Generated by Cython 3.1.6
Yellow lines hint at Python interaction.
Click on a line that starts with a "+" to see the C code that Cython generated for it.
Raw output: jsonl_reader.cpp
+001: """Cython wrapper for reading JSONL files."""
__pyx_t_2 = __Pyx_PyDict_NewPresized(0); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 1, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_2); if (PyDict_SetItem(__pyx_mstate_global->__pyx_d, __pyx_mstate_global->__pyx_n_u_test, __pyx_t_2) < (0)) __PYX_ERR(0, 1, __pyx_L1_error) __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
002:
003: # distutils: language = c++
004: # cython: language_level=3
005: # cython: nonecheck=False
006: # cython: cdivision=True
007: # cython: boundscheck=False
008: # cython: wraparound=False
009: # cython: infer_types=True
010:
011: from libc.stdint cimport uint8_t, int64_t
012: from libcpp.string cimport string
013: from libcpp.vector cimport vector
014: from cpython.buffer cimport PyBUF_CONTIG_RO, PyObject_GetBuffer, PyBuffer_Release, Py_buffer
+015: import json
__pyx_t_2 = __Pyx_ImportDottedModule(__pyx_mstate_global->__pyx_n_u_json, NULL); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 15, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_2); if (PyDict_SetItem(__pyx_mstate_global->__pyx_d, __pyx_mstate_global->__pyx_n_u_json, __pyx_t_2) < (0)) __PYX_ERR(0, 15, __pyx_L1_error) __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
016: from cpython.ref cimport PyObject, Py_INCREF, Py_DECREF
017: from cpython.exc cimport PyErr_Occurred
018: from cpython.list cimport PyList_Append
019: from cpython.exc cimport PyErr_Clear
020: from cpython.bytes cimport PyBytes_FromStringAndSize
021:
022: # Internal fast array parser (no runtime deps). Parses a JSON array encoded
023: # in UTF-8 bytes into Python lists. Objects found inside arrays are returned
024: # as raw bytes; strings are unescaped to Python str; numbers become int/float;
025: # null -> None, true/false -> bool.
026:
027:
028: # (removed unused fast whitespace skip function)
029:
030:
+031: def _parse_array_from_bytes(bytes b):
/* Python wrapper */
static PyObject *__pyx_pw_4rugo_5jsonl_1_parse_array_from_bytes(PyObject *__pyx_self,
#if CYTHON_METH_FASTCALL
PyObject *const *__pyx_args, Py_ssize_t __pyx_nargs, PyObject *__pyx_kwds
#else
PyObject *__pyx_args, PyObject *__pyx_kwds
#endif
); /*proto*/
static PyMethodDef __pyx_mdef_4rugo_5jsonl_1_parse_array_from_bytes = {"_parse_array_from_bytes", (PyCFunction)(void(*)(void))(__Pyx_PyCFunction_FastCallWithKeywords)__pyx_pw_4rugo_5jsonl_1_parse_array_from_bytes, __Pyx_METH_FASTCALL|METH_KEYWORDS, 0};
static PyObject *__pyx_pw_4rugo_5jsonl_1_parse_array_from_bytes(PyObject *__pyx_self,
#if CYTHON_METH_FASTCALL
PyObject *const *__pyx_args, Py_ssize_t __pyx_nargs, PyObject *__pyx_kwds
#else
PyObject *__pyx_args, PyObject *__pyx_kwds
#endif
) {
PyObject *__pyx_v_b = 0;
#if !CYTHON_METH_FASTCALL
CYTHON_UNUSED Py_ssize_t __pyx_nargs;
#endif
CYTHON_UNUSED PyObject *const *__pyx_kwvalues;
PyObject *__pyx_r = 0;
__Pyx_RefNannyDeclarations
__Pyx_RefNannySetupContext("_parse_array_from_bytes (wrapper)", 0);
#if !CYTHON_METH_FASTCALL
#if CYTHON_ASSUME_SAFE_SIZE
__pyx_nargs = PyTuple_GET_SIZE(__pyx_args);
#else
__pyx_nargs = PyTuple_Size(__pyx_args); if (unlikely(__pyx_nargs < 0)) return NULL;
#endif
#endif
__pyx_kwvalues = __Pyx_KwValues_FASTCALL(__pyx_args, __pyx_nargs);
{
PyObject ** const __pyx_pyargnames[] = {&__pyx_mstate_global->__pyx_n_u_b,0};
PyObject* values[1] = {0};
const Py_ssize_t __pyx_kwds_len = (__pyx_kwds) ? __Pyx_NumKwargs_FASTCALL(__pyx_kwds) : 0;
if (unlikely(__pyx_kwds_len) < 0) __PYX_ERR(0, 31, __pyx_L3_error)
if (__pyx_kwds_len > 0) {
switch (__pyx_nargs) {
case 1:
values[0] = __Pyx_ArgRef_FASTCALL(__pyx_args, 0);
if (!CYTHON_ASSUME_SAFE_MACROS && unlikely(!values[0])) __PYX_ERR(0, 31, __pyx_L3_error)
CYTHON_FALLTHROUGH;
case 0: break;
default: goto __pyx_L5_argtuple_error;
}
const Py_ssize_t kwd_pos_args = __pyx_nargs;
if (__Pyx_ParseKeywords(__pyx_kwds, __pyx_kwvalues, __pyx_pyargnames, 0, values, kwd_pos_args, __pyx_kwds_len, "_parse_array_from_bytes", 0) < (0)) __PYX_ERR(0, 31, __pyx_L3_error)
for (Py_ssize_t i = __pyx_nargs; i < 1; i++) {
if (unlikely(!values[i])) { __Pyx_RaiseArgtupleInvalid("_parse_array_from_bytes", 1, 1, 1, i); __PYX_ERR(0, 31, __pyx_L3_error) }
}
} else if (unlikely(__pyx_nargs != 1)) {
goto __pyx_L5_argtuple_error;
} else {
values[0] = __Pyx_ArgRef_FASTCALL(__pyx_args, 0);
if (!CYTHON_ASSUME_SAFE_MACROS && unlikely(!values[0])) __PYX_ERR(0, 31, __pyx_L3_error)
}
__pyx_v_b = ((PyObject*)values[0]);
}
goto __pyx_L6_skip;
__pyx_L5_argtuple_error:;
__Pyx_RaiseArgtupleInvalid("_parse_array_from_bytes", 1, 1, 1, __pyx_nargs); __PYX_ERR(0, 31, __pyx_L3_error)
__pyx_L6_skip:;
goto __pyx_L4_argument_unpacking_done;
__pyx_L3_error:;
for (Py_ssize_t __pyx_temp=0; __pyx_temp < (Py_ssize_t)(sizeof(values)/sizeof(values[0])); ++__pyx_temp) {
Py_XDECREF(values[__pyx_temp]);
}
__Pyx_AddTraceback("rugo.jsonl._parse_array_from_bytes", __pyx_clineno, __pyx_lineno, __pyx_filename);
__Pyx_RefNannyFinishContext();
return NULL;
__pyx_L4_argument_unpacking_done:;
if (unlikely(!__Pyx_ArgTypeTest(((PyObject *)__pyx_v_b), (&PyBytes_Type), 1, "b", 1))) __PYX_ERR(0, 31, __pyx_L1_error)
__pyx_r = __pyx_pf_4rugo_5jsonl__parse_array_from_bytes(__pyx_self, __pyx_v_b);
int __pyx_lineno = 0;
const char *__pyx_filename = NULL;
int __pyx_clineno = 0;
/* function exit code */
goto __pyx_L0;
__pyx_L1_error:;
__pyx_r = NULL;
for (Py_ssize_t __pyx_temp=0; __pyx_temp < (Py_ssize_t)(sizeof(values)/sizeof(values[0])); ++__pyx_temp) {
Py_XDECREF(values[__pyx_temp]);
}
goto __pyx_L7_cleaned_up;
__pyx_L0:;
for (Py_ssize_t __pyx_temp=0; __pyx_temp < (Py_ssize_t)(sizeof(values)/sizeof(values[0])); ++__pyx_temp) {
Py_XDECREF(values[__pyx_temp]);
}
__pyx_L7_cleaned_up:;
__Pyx_RefNannyFinishContext();
return __pyx_r;
}
/* … */
static PyObject *__pyx_pf_4rugo_5jsonl__parse_array_from_bytes(CYTHON_UNUSED PyObject *__pyx_self, PyObject *__pyx_v_b) {
struct __pyx_obj_4rugo_5jsonl___pyx_scope_struct___parse_array_from_bytes *__pyx_cur_scope;
PyObject *__pyx_r = NULL;
__pyx_cur_scope = (struct __pyx_obj_4rugo_5jsonl___pyx_scope_struct___parse_array_from_bytes *)__pyx_tp_new_4rugo_5jsonl___pyx_scope_struct___parse_array_from_bytes(__pyx_mstate_global->__pyx_ptype_4rugo_5jsonl___pyx_scope_struct___parse_array_from_bytes, __pyx_mstate_global->__pyx_empty_tuple, NULL);
if (unlikely(!__pyx_cur_scope)) {
__pyx_cur_scope = ((struct __pyx_obj_4rugo_5jsonl___pyx_scope_struct___parse_array_from_bytes *)Py_None);
__Pyx_INCREF(Py_None);
__PYX_ERR(0, 31, __pyx_L1_error)
} else {
__Pyx_GOTREF((PyObject *)__pyx_cur_scope);
}
__pyx_cur_scope->__pyx_v_b = __pyx_v_b;
__Pyx_INCREF(__pyx_cur_scope->__pyx_v_b);
__Pyx_GIVEREF(__pyx_cur_scope->__pyx_v_b);
/* … */
/* function exit code */
__pyx_L1_error:;
__Pyx_XDECREF(__pyx_t_1);
__Pyx_XDECREF(__pyx_t_6);
__Pyx_XDECREF(__pyx_t_7);
__Pyx_AddTraceback("rugo.jsonl._parse_array_from_bytes", __pyx_clineno, __pyx_lineno, __pyx_filename);
__pyx_r = NULL;
__pyx_L0:;
__Pyx_DECREF((PyObject *)__pyx_cur_scope);
__Pyx_XGIVEREF(__pyx_r);
__Pyx_RefNannyFinishContext();
return __pyx_r;
}
/* … */
__pyx_t_2 = __Pyx_CyFunction_New(&__pyx_mdef_4rugo_5jsonl_1_parse_array_from_bytes, 0, __pyx_mstate_global->__pyx_n_u_parse_array_from_bytes, NULL, __pyx_mstate_global->__pyx_n_u_rugo_jsonl, __pyx_mstate_global->__pyx_d, ((PyObject *)__pyx_mstate_global->__pyx_codeobj_tab[4])); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 31, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_2);
if (PyDict_SetItem(__pyx_mstate_global->__pyx_d, __pyx_mstate_global->__pyx_n_u_parse_array_from_bytes, __pyx_t_2) < (0)) __PYX_ERR(0, 31, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
/* … */
struct __pyx_obj_4rugo_5jsonl___pyx_scope_struct___parse_array_from_bytes {
PyObject_HEAD
PyObject *__pyx_v_b;
Py_ssize_t __pyx_v_i;
Py_ssize_t __pyx_v_n;
PyObject *__pyx_v_parse_value;
};
+032: cdef Py_ssize_t i = 0
__pyx_cur_scope->__pyx_v_i = 0;
+033: cdef Py_ssize_t n = len(b)
__pyx_t_1 = __pyx_cur_scope->__pyx_v_b; __Pyx_INCREF(__pyx_t_1); if (unlikely(__pyx_t_1 == Py_None)) { PyErr_SetString(PyExc_TypeError, "object of type 'NoneType' has no len()"); __PYX_ERR(0, 33, __pyx_L1_error) } __pyx_t_2 = __Pyx_PyBytes_GET_SIZE(__pyx_t_1); if (unlikely(__pyx_t_2 == ((Py_ssize_t)-1))) __PYX_ERR(0, 33, __pyx_L1_error) __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0; __pyx_cur_scope->__pyx_v_n = __pyx_t_2;
034:
+035: def parse_value():
/* Python wrapper */
static PyObject *__pyx_pw_4rugo_5jsonl_23_parse_array_from_bytes_1parse_value(PyObject *__pyx_self, CYTHON_UNUSED PyObject *unused); /*proto*/
static PyMethodDef __pyx_mdef_4rugo_5jsonl_23_parse_array_from_bytes_1parse_value = {"parse_value", (PyCFunction)__pyx_pw_4rugo_5jsonl_23_parse_array_from_bytes_1parse_value, METH_NOARGS, 0};
static PyObject *__pyx_pw_4rugo_5jsonl_23_parse_array_from_bytes_1parse_value(PyObject *__pyx_self, CYTHON_UNUSED PyObject *unused) {
CYTHON_UNUSED PyObject *const *__pyx_kwvalues;
PyObject *__pyx_r = 0;
__Pyx_RefNannyDeclarations
__Pyx_RefNannySetupContext("parse_value (wrapper)", 0);
__pyx_kwvalues = __Pyx_KwValues_VARARGS(__pyx_args, __pyx_nargs);
__pyx_r = __pyx_pf_4rugo_5jsonl_23_parse_array_from_bytes_parse_value(__pyx_self);
/* function exit code */
__Pyx_RefNannyFinishContext();
return __pyx_r;
}
static PyObject *__pyx_pf_4rugo_5jsonl_23_parse_array_from_bytes_parse_value(PyObject *__pyx_self) {
struct __pyx_obj_4rugo_5jsonl___pyx_scope_struct___parse_array_from_bytes *__pyx_cur_scope;
struct __pyx_obj_4rugo_5jsonl___pyx_scope_struct___parse_array_from_bytes *__pyx_outer_scope;
PyObject *__pyx_v_c = NULL;
PyObject *__pyx_v_chars = NULL;
PyObject *__pyx_v_ch = NULL;
PyObject *__pyx_v_esc = NULL;
PyObject *__pyx_v_hex_s = NULL;
PyObject *__pyx_v_cp = NULL;
Py_ssize_t __pyx_v_start;
PyObject *__pyx_v_s = NULL;
PyObject *__pyx_v_res = NULL;
PyObject *__pyx_v_val = NULL;
PyObject *__pyx_v_depth = NULL;
PyObject *__pyx_r = NULL;
__pyx_outer_scope = (struct __pyx_obj_4rugo_5jsonl___pyx_scope_struct___parse_array_from_bytes *) __Pyx_CyFunction_GetClosure(__pyx_self);
__pyx_cur_scope = __pyx_outer_scope;
/* … */
/* function exit code */
__pyx_L1_error:;
__Pyx_XDECREF(__pyx_t_3);
__Pyx_XDECREF(__pyx_t_5);
__Pyx_XDECREF(__pyx_t_6);
__Pyx_XDECREF(__pyx_t_13);
__Pyx_XDECREF(__pyx_t_15);
__Pyx_AddTraceback("rugo.jsonl._parse_array_from_bytes.parse_value", __pyx_clineno, __pyx_lineno, __pyx_filename);
__pyx_r = NULL;
__pyx_L0:;
__Pyx_XDECREF(__pyx_v_c);
__Pyx_XDECREF(__pyx_v_chars);
__Pyx_XDECREF(__pyx_v_ch);
__Pyx_XDECREF(__pyx_v_esc);
__Pyx_XDECREF(__pyx_v_hex_s);
__Pyx_XDECREF(__pyx_v_cp);
__Pyx_XDECREF(__pyx_v_s);
__Pyx_XDECREF(__pyx_v_res);
__Pyx_XDECREF(__pyx_v_val);
__Pyx_XDECREF(__pyx_v_depth);
__Pyx_XGIVEREF(__pyx_r);
__Pyx_RefNannyFinishContext();
return __pyx_r;
}
/* … */
__pyx_t_1 = __Pyx_CyFunction_New(&__pyx_mdef_4rugo_5jsonl_23_parse_array_from_bytes_1parse_value, 0, __pyx_mstate_global->__pyx_n_u_parse_array_from_bytes_locals_p, ((PyObject*)__pyx_cur_scope), __pyx_mstate_global->__pyx_n_u_rugo_jsonl, __pyx_mstate_global->__pyx_d, ((PyObject *)__pyx_mstate_global->__pyx_codeobj_tab[0])); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 35, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_1);
__Pyx_GIVEREF(__pyx_t_1);
__pyx_cur_scope->__pyx_v_parse_value = __pyx_t_1;
__pyx_t_1 = 0;
036: nonlocal i
037: # skip whitespace
+038: while i < n and b[i] in (32,9,10,13):
while (1) {
__pyx_t_2 = (__pyx_cur_scope->__pyx_v_i < __pyx_cur_scope->__pyx_v_n);
if (__pyx_t_2) {
} else {
__pyx_t_1 = __pyx_t_2;
goto __pyx_L5_bool_binop_done;
}
if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 38, __pyx_L1_error) }
__pyx_t_3 = __Pyx_GetItemInt(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, Py_ssize_t, 1, PyLong_FromSsize_t, 0, 0, 0, 1); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 38, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_3);
__pyx_t_4 = (__Pyx_PyLong_BoolEqObjC(__pyx_t_3, __pyx_mstate_global->__pyx_int_32, 32, 0)); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 38, __pyx_L1_error)
if (!__pyx_t_4) {
} else {
__pyx_t_2 = __pyx_t_4;
goto __pyx_L7_bool_binop_done;
}
__pyx_t_4 = (__Pyx_PyLong_BoolEqObjC(__pyx_t_3, __pyx_mstate_global->__pyx_int_9, 9, 0)); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 38, __pyx_L1_error)
if (!__pyx_t_4) {
} else {
__pyx_t_2 = __pyx_t_4;
goto __pyx_L7_bool_binop_done;
}
__pyx_t_4 = (__Pyx_PyLong_BoolEqObjC(__pyx_t_3, __pyx_mstate_global->__pyx_int_10, 10, 0)); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 38, __pyx_L1_error)
if (!__pyx_t_4) {
} else {
__pyx_t_2 = __pyx_t_4;
goto __pyx_L7_bool_binop_done;
}
__pyx_t_4 = (__Pyx_PyLong_BoolEqObjC(__pyx_t_3, __pyx_mstate_global->__pyx_int_13, 13, 0)); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 38, __pyx_L1_error)
__pyx_t_2 = __pyx_t_4;
__pyx_L7_bool_binop_done:;
__Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
__pyx_t_4 = __pyx_t_2;
__pyx_t_1 = __pyx_t_4;
__pyx_L5_bool_binop_done:;
if (!__pyx_t_1) break;
+039: i += 1
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 1); }
+040: if i >= n:
__pyx_t_1 = (__pyx_cur_scope->__pyx_v_i >= __pyx_cur_scope->__pyx_v_n);
if (unlikely(__pyx_t_1)) {
/* … */
}
+041: raise ValueError('unexpected end')
__pyx_t_5 = NULL;
__Pyx_INCREF(__pyx_builtin_ValueError);
__pyx_t_6 = __pyx_builtin_ValueError;
__pyx_t_7 = 1;
{
PyObject *__pyx_callargs[2] = {__pyx_t_5, __pyx_mstate_global->__pyx_kp_u_unexpected_end};
__pyx_t_3 = __Pyx_PyObject_FastCall(__pyx_t_6, __pyx_callargs+__pyx_t_7, (2-__pyx_t_7) | (__pyx_t_7*__Pyx_PY_VECTORCALL_ARGUMENTS_OFFSET));
__Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
__Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 41, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_3);
}
__Pyx_Raise(__pyx_t_3, 0, 0, 0);
__Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
__PYX_ERR(0, 41, __pyx_L1_error)
+042: c = b[i]
if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 42, __pyx_L1_error) }
__pyx_t_3 = __Pyx_GetItemInt(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, Py_ssize_t, 1, PyLong_FromSsize_t, 0, 0, 0, 1); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 42, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_3);
__pyx_v_c = __pyx_t_3;
__pyx_t_3 = 0;
043: # string
+044: if c == 34: # '"'
__pyx_t_1 = (__Pyx_PyLong_BoolEqObjC(__pyx_v_c, __pyx_mstate_global->__pyx_int_34, 34, 0)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 44, __pyx_L1_error) if (__pyx_t_1) { /* … */ }
+045: i += 1
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 1);
+046: chars = []
__pyx_t_3 = PyList_New(0); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 46, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_3); __pyx_v_chars = ((PyObject*)__pyx_t_3); __pyx_t_3 = 0;
+047: while i < n:
while (1) {
__pyx_t_1 = (__pyx_cur_scope->__pyx_v_i < __pyx_cur_scope->__pyx_v_n);
if (!__pyx_t_1) break;
+048: ch = b[i]
if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 48, __pyx_L1_error) }
__pyx_t_3 = __Pyx_GetItemInt(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, Py_ssize_t, 1, PyLong_FromSsize_t, 0, 0, 0, 1); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 48, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_3);
__Pyx_XDECREF_SET(__pyx_v_ch, __pyx_t_3);
__pyx_t_3 = 0;
+049: if ch == 34:
__pyx_t_1 = (__Pyx_PyLong_BoolEqObjC(__pyx_v_ch, __pyx_mstate_global->__pyx_int_34, 34, 0)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 49, __pyx_L1_error) if (__pyx_t_1) { /* … */ }
+050: i += 1
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 1);
+051: return ''.join(chars)
__Pyx_XDECREF(__pyx_r); __pyx_t_3 = PyUnicode_Join(__pyx_mstate_global->__pyx_kp_u_, __pyx_v_chars); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 51, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_3); __pyx_r = __pyx_t_3; __pyx_t_3 = 0; goto __pyx_L0;
+052: if ch == 92: # backslash
__pyx_t_1 = (__Pyx_PyLong_BoolEqObjC(__pyx_v_ch, __pyx_mstate_global->__pyx_int_92, 92, 0)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 52, __pyx_L1_error) if (__pyx_t_1) { /* … */ goto __pyx_L16; }
+053: i += 1
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 1);
+054: if i >= n:
__pyx_t_1 = (__pyx_cur_scope->__pyx_v_i >= __pyx_cur_scope->__pyx_v_n);
if (unlikely(__pyx_t_1)) {
/* … */
}
+055: raise ValueError('unterminated escape')
__pyx_t_6 = NULL;
__Pyx_INCREF(__pyx_builtin_ValueError);
__pyx_t_5 = __pyx_builtin_ValueError;
__pyx_t_7 = 1;
{
PyObject *__pyx_callargs[2] = {__pyx_t_6, __pyx_mstate_global->__pyx_kp_u_unterminated_escape};
__pyx_t_3 = __Pyx_PyObject_FastCall(__pyx_t_5, __pyx_callargs+__pyx_t_7, (2-__pyx_t_7) | (__pyx_t_7*__Pyx_PY_VECTORCALL_ARGUMENTS_OFFSET));
__Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
__Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 55, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_3);
}
__Pyx_Raise(__pyx_t_3, 0, 0, 0);
__Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
__PYX_ERR(0, 55, __pyx_L1_error)
+056: esc = b[i]
if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 56, __pyx_L1_error) }
__pyx_t_3 = __Pyx_GetItemInt(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, Py_ssize_t, 1, PyLong_FromSsize_t, 0, 0, 0, 1); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 56, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_3);
__Pyx_XDECREF_SET(__pyx_v_esc, __pyx_t_3);
__pyx_t_3 = 0;
+057: i += 1
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 1);
+058: if esc == 34: chars.append('"')
__pyx_t_1 = (__Pyx_PyLong_BoolEqObjC(__pyx_v_esc, __pyx_mstate_global->__pyx_int_34, 34, 0)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 58, __pyx_L1_error) if (__pyx_t_1) { __pyx_t_8 = __Pyx_PyList_Append(__pyx_v_chars, __pyx_mstate_global->__pyx_kp_u__2); if (unlikely(__pyx_t_8 == ((int)-1))) __PYX_ERR(0, 58, __pyx_L1_error) goto __pyx_L18; }
+059: elif esc == 92: chars.append('\\')
__pyx_t_1 = (__Pyx_PyLong_BoolEqObjC(__pyx_v_esc, __pyx_mstate_global->__pyx_int_92, 92, 0)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 59, __pyx_L1_error) if (__pyx_t_1) { __pyx_t_8 = __Pyx_PyList_Append(__pyx_v_chars, __pyx_mstate_global->__pyx_kp_u__3); if (unlikely(__pyx_t_8 == ((int)-1))) __PYX_ERR(0, 59, __pyx_L1_error) goto __pyx_L18; }
+060: elif esc == 47: chars.append('/')
__pyx_t_1 = (__Pyx_PyLong_BoolEqObjC(__pyx_v_esc, __pyx_mstate_global->__pyx_int_47, 47, 0)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 60, __pyx_L1_error) if (__pyx_t_1) { __pyx_t_8 = __Pyx_PyList_Append(__pyx_v_chars, __pyx_mstate_global->__pyx_kp_u__4); if (unlikely(__pyx_t_8 == ((int)-1))) __PYX_ERR(0, 60, __pyx_L1_error) goto __pyx_L18; }
+061: elif esc == 98: chars.append('\b')
__pyx_t_1 = (__Pyx_PyLong_BoolEqObjC(__pyx_v_esc, __pyx_mstate_global->__pyx_int_98, 98, 0)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 61, __pyx_L1_error) if (__pyx_t_1) { __pyx_t_8 = __Pyx_PyList_Append(__pyx_v_chars, __pyx_mstate_global->__pyx_kp_u__5); if (unlikely(__pyx_t_8 == ((int)-1))) __PYX_ERR(0, 61, __pyx_L1_error) goto __pyx_L18; }
+062: elif esc == 102: chars.append('\f')
__pyx_t_1 = (__Pyx_PyLong_BoolEqObjC(__pyx_v_esc, __pyx_mstate_global->__pyx_int_102, 0x66, 0)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 62, __pyx_L1_error) if (__pyx_t_1) { __pyx_t_8 = __Pyx_PyList_Append(__pyx_v_chars, __pyx_mstate_global->__pyx_kp_u__6); if (unlikely(__pyx_t_8 == ((int)-1))) __PYX_ERR(0, 62, __pyx_L1_error) goto __pyx_L18; }
+063: elif esc == 110: chars.append('\n')
__pyx_t_1 = (__Pyx_PyLong_BoolEqObjC(__pyx_v_esc, __pyx_mstate_global->__pyx_int_110, 0x6E, 0)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 63, __pyx_L1_error) if (__pyx_t_1) { __pyx_t_8 = __Pyx_PyList_Append(__pyx_v_chars, __pyx_mstate_global->__pyx_kp_u__7); if (unlikely(__pyx_t_8 == ((int)-1))) __PYX_ERR(0, 63, __pyx_L1_error) goto __pyx_L18; }
+064: elif esc == 114: chars.append('\r')
__pyx_t_1 = (__Pyx_PyLong_BoolEqObjC(__pyx_v_esc, __pyx_mstate_global->__pyx_int_114, 0x72, 0)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 64, __pyx_L1_error) if (__pyx_t_1) { __pyx_t_8 = __Pyx_PyList_Append(__pyx_v_chars, __pyx_mstate_global->__pyx_kp_u__8); if (unlikely(__pyx_t_8 == ((int)-1))) __PYX_ERR(0, 64, __pyx_L1_error) goto __pyx_L18; }
+065: elif esc == 116: chars.append('\t')
__pyx_t_1 = (__Pyx_PyLong_BoolEqObjC(__pyx_v_esc, __pyx_mstate_global->__pyx_int_116, 0x74, 0)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 65, __pyx_L1_error) if (__pyx_t_1) { __pyx_t_8 = __Pyx_PyList_Append(__pyx_v_chars, __pyx_mstate_global->__pyx_kp_u__9); if (unlikely(__pyx_t_8 == ((int)-1))) __PYX_ERR(0, 65, __pyx_L1_error) goto __pyx_L18; }
+066: elif esc == 117:
__pyx_t_1 = (__Pyx_PyLong_BoolEqObjC(__pyx_v_esc, __pyx_mstate_global->__pyx_int_117, 0x75, 0)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 66, __pyx_L1_error) if (__pyx_t_1) { /* … */ goto __pyx_L18; }
067: # unicode escape \uXXXX
+068: if i + 4 <= n:
__pyx_t_1 = ((__pyx_cur_scope->__pyx_v_i + 4) <= __pyx_cur_scope->__pyx_v_n);
if (likely(__pyx_t_1)) {
/* … */
goto __pyx_L19;
}
+069: hex_s = b[i:i+4].decode('ascii')
if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 69, __pyx_L1_error) }
if (unlikely(__pyx_cur_scope->__pyx_v_b == Py_None)) {
PyErr_SetString(PyExc_TypeError, "'NoneType' object is not subscriptable");
__PYX_ERR(0, 69, __pyx_L1_error)
}
__pyx_t_3 = __Pyx_decode_bytes(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, (__pyx_cur_scope->__pyx_v_i + 4), NULL, NULL, PyUnicode_DecodeASCII); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 69, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_3);
__Pyx_XDECREF_SET(__pyx_v_hex_s, ((PyObject*)__pyx_t_3));
__pyx_t_3 = 0;
+070: try:
{
/*try:*/ {
/* … */
}
__Pyx_XDECREF(__pyx_t_9); __pyx_t_9 = 0;
__Pyx_XDECREF(__pyx_t_10); __pyx_t_10 = 0;
__Pyx_XDECREF(__pyx_t_11); __pyx_t_11 = 0;
goto __pyx_L27_try_end;
__pyx_L20_error:;
__Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;
__Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
__Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
/* … */
__pyx_L22_except_error:;
__Pyx_XGIVEREF(__pyx_t_9);
__Pyx_XGIVEREF(__pyx_t_10);
__Pyx_XGIVEREF(__pyx_t_11);
__Pyx_ExceptionReset(__pyx_t_9, __pyx_t_10, __pyx_t_11);
goto __pyx_L1_error;
__pyx_L21_exception_handled:;
__Pyx_XGIVEREF(__pyx_t_9);
__Pyx_XGIVEREF(__pyx_t_10);
__Pyx_XGIVEREF(__pyx_t_11);
__Pyx_ExceptionReset(__pyx_t_9, __pyx_t_10, __pyx_t_11);
__pyx_L27_try_end:;
}
+071: cp = int(hex_s, 16)
__pyx_t_5 = NULL;
__Pyx_INCREF((PyObject *)(&PyLong_Type));
__pyx_t_6 = ((PyObject *)(&PyLong_Type));
__pyx_t_7 = 1;
{
PyObject *__pyx_callargs[3] = {__pyx_t_5, __pyx_v_hex_s, __pyx_mstate_global->__pyx_int_16};
__pyx_t_3 = __Pyx_PyObject_FastCall(__pyx_t_6, __pyx_callargs+__pyx_t_7, (3-__pyx_t_7) | (__pyx_t_7*__Pyx_PY_VECTORCALL_ARGUMENTS_OFFSET));
__Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
__Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 71, __pyx_L20_error)
__Pyx_GOTREF(__pyx_t_3);
}
__Pyx_XDECREF_SET(__pyx_v_cp, ((PyObject*)__pyx_t_3));
__pyx_t_3 = 0;
+072: chars.append(chr(cp))
__pyx_t_12 = __Pyx_PyLong_As_int(__pyx_v_cp); if (unlikely((__pyx_t_12 == (int)-1) && PyErr_Occurred())) __PYX_ERR(0, 72, __pyx_L20_error) __pyx_t_3 = PyUnicode_FromOrdinal(__pyx_t_12); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 72, __pyx_L20_error) __Pyx_GOTREF(__pyx_t_3); __pyx_t_8 = __Pyx_PyList_Append(__pyx_v_chars, __pyx_t_3); if (unlikely(__pyx_t_8 == ((int)-1))) __PYX_ERR(0, 72, __pyx_L20_error) __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
+073: except Exception:
__pyx_t_12 = __Pyx_PyErr_ExceptionMatches(((PyObject *)(((PyTypeObject*)PyExc_Exception)))); if (__pyx_t_12) { __Pyx_AddTraceback("rugo.jsonl._parse_array_from_bytes.parse_value", __pyx_clineno, __pyx_lineno, __pyx_filename); if (__Pyx_GetException(&__pyx_t_3, &__pyx_t_6, &__pyx_t_5) < 0) __PYX_ERR(0, 73, __pyx_L22_except_error) __Pyx_XGOTREF(__pyx_t_3); __Pyx_XGOTREF(__pyx_t_6); __Pyx_XGOTREF(__pyx_t_5);
+074: chars.append('\\u' + hex_s)
__pyx_t_13 = __Pyx_PyUnicode_Concat(__pyx_mstate_global->__pyx_kp_u_u, __pyx_v_hex_s); if (unlikely(!__pyx_t_13)) __PYX_ERR(0, 74, __pyx_L22_except_error) __Pyx_GOTREF(__pyx_t_13); __pyx_t_8 = __Pyx_PyList_Append(__pyx_v_chars, __pyx_t_13); if (unlikely(__pyx_t_8 == ((int)-1))) __PYX_ERR(0, 74, __pyx_L22_except_error) __Pyx_DECREF(__pyx_t_13); __pyx_t_13 = 0; __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0; __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0; __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0; goto __pyx_L21_exception_handled; } goto __pyx_L22_except_error;
+075: i += 4
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 4);
076: else:
+077: raise ValueError('invalid unicode escape')
/*else*/ {
__pyx_t_6 = NULL;
__Pyx_INCREF(__pyx_builtin_ValueError);
__pyx_t_3 = __pyx_builtin_ValueError;
__pyx_t_7 = 1;
{
PyObject *__pyx_callargs[2] = {__pyx_t_6, __pyx_mstate_global->__pyx_kp_u_invalid_unicode_escape};
__pyx_t_5 = __Pyx_PyObject_FastCall(__pyx_t_3, __pyx_callargs+__pyx_t_7, (2-__pyx_t_7) | (__pyx_t_7*__Pyx_PY_VECTORCALL_ARGUMENTS_OFFSET));
__Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
__Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 77, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_5);
}
__Pyx_Raise(__pyx_t_5, 0, 0, 0);
__Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
__PYX_ERR(0, 77, __pyx_L1_error)
}
__pyx_L19:;
078: else:
079: # unknown escape, keep char
+080: chars.append(chr(esc))
/*else*/ {
__pyx_t_12 = __Pyx_PyLong_As_int(__pyx_v_esc); if (unlikely((__pyx_t_12 == (int)-1) && PyErr_Occurred())) __PYX_ERR(0, 80, __pyx_L1_error)
__pyx_t_5 = PyUnicode_FromOrdinal(__pyx_t_12); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 80, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_5);
__pyx_t_8 = __Pyx_PyList_Append(__pyx_v_chars, __pyx_t_5); if (unlikely(__pyx_t_8 == ((int)-1))) __PYX_ERR(0, 80, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
}
__pyx_L18:;
081: else:
082: # append raw utf-8 byte; accumulate then decode at end
083: # to support multi-byte UTF-8 sequences, collect bytes
+084: start = i
/*else*/ {
__pyx_v_start = __pyx_cur_scope->__pyx_v_i;
085: # collect consecutive non-escape non-quote bytes
+086: while i < n and b[i] != 34 and b[i] != 92:
while (1) {
__pyx_t_4 = (__pyx_cur_scope->__pyx_v_i < __pyx_cur_scope->__pyx_v_n);
if (__pyx_t_4) {
} else {
__pyx_t_1 = __pyx_t_4;
goto __pyx_L32_bool_binop_done;
}
if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 86, __pyx_L1_error) }
__pyx_t_5 = __Pyx_GetItemInt(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, Py_ssize_t, 1, PyLong_FromSsize_t, 0, 0, 0, 1); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 86, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_5);
__pyx_t_4 = (__Pyx_PyLong_BoolNeObjC(__pyx_t_5, __pyx_mstate_global->__pyx_int_34, 34, 0)); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 86, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
if (__pyx_t_4) {
} else {
__pyx_t_1 = __pyx_t_4;
goto __pyx_L32_bool_binop_done;
}
if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 86, __pyx_L1_error) }
__pyx_t_5 = __Pyx_GetItemInt(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, Py_ssize_t, 1, PyLong_FromSsize_t, 0, 0, 0, 1); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 86, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_5);
__pyx_t_4 = (__Pyx_PyLong_BoolNeObjC(__pyx_t_5, __pyx_mstate_global->__pyx_int_92, 92, 0)); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 86, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
__pyx_t_1 = __pyx_t_4;
__pyx_L32_bool_binop_done:;
if (!__pyx_t_1) break;
+087: i += 1
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 1);
}
088: # decode slice
+089: chars.append(b[start:i].decode('utf-8'))
if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 89, __pyx_L1_error) }
if (unlikely(__pyx_cur_scope->__pyx_v_b == Py_None)) {
PyErr_SetString(PyExc_TypeError, "'NoneType' object is not subscriptable");
__PYX_ERR(0, 89, __pyx_L1_error)
}
__pyx_t_5 = __Pyx_decode_bytes(__pyx_cur_scope->__pyx_v_b, __pyx_v_start, __pyx_cur_scope->__pyx_v_i, NULL, NULL, PyUnicode_DecodeUTF8); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 89, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_5);
__pyx_t_8 = __Pyx_PyList_Append(__pyx_v_chars, __pyx_t_5); if (unlikely(__pyx_t_8 == ((int)-1))) __PYX_ERR(0, 89, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
}
__pyx_L16:;
}
+090: raise ValueError('unterminated string')
__pyx_t_3 = NULL;
__Pyx_INCREF(__pyx_builtin_ValueError);
__pyx_t_6 = __pyx_builtin_ValueError;
__pyx_t_7 = 1;
{
PyObject *__pyx_callargs[2] = {__pyx_t_3, __pyx_mstate_global->__pyx_kp_u_unterminated_string};
__pyx_t_5 = __Pyx_PyObject_FastCall(__pyx_t_6, __pyx_callargs+__pyx_t_7, (2-__pyx_t_7) | (__pyx_t_7*__Pyx_PY_VECTORCALL_ARGUMENTS_OFFSET));
__Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;
__Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 90, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_5);
}
__Pyx_Raise(__pyx_t_5, 0, 0, 0);
__Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
__PYX_ERR(0, 90, __pyx_L1_error)
091:
092: # null
+093: if c == 110 and i + 4 <= n and b[i:i+4] == b'null':
__pyx_t_4 = (__Pyx_PyLong_BoolEqObjC(__pyx_v_c, __pyx_mstate_global->__pyx_int_110, 0x6E, 0)); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 93, __pyx_L1_error) if (__pyx_t_4) { } else { __pyx_t_1 = __pyx_t_4; goto __pyx_L36_bool_binop_done; } __pyx_t_4 = ((__pyx_cur_scope->__pyx_v_i + 4) <= __pyx_cur_scope->__pyx_v_n); if (__pyx_t_4) { } else { __pyx_t_1 = __pyx_t_4; goto __pyx_L36_bool_binop_done; } if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 93, __pyx_L1_error) } if (unlikely(__pyx_cur_scope->__pyx_v_b == Py_None)) { PyErr_SetString(PyExc_TypeError, "'NoneType' object is not subscriptable"); __PYX_ERR(0, 93, __pyx_L1_error) } __pyx_t_5 = PySequence_GetSlice(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, (__pyx_cur_scope->__pyx_v_i + 4)); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 93, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_5); __pyx_t_4 = (__Pyx_PyBytes_Equals(__pyx_t_5, __pyx_mstate_global->__pyx_n_b_null, Py_EQ)); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 93, __pyx_L1_error) __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; __pyx_t_1 = __pyx_t_4; __pyx_L36_bool_binop_done:; if (__pyx_t_1) { /* … */ }
+094: i += 4
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 4);
+095: return None
__Pyx_XDECREF(__pyx_r); __pyx_r = Py_None; __Pyx_INCREF(Py_None); goto __pyx_L0;
096:
097: # true/false
+098: if c == 116 and i + 4 <= n and b[i:i+4] == b'true':
__pyx_t_4 = (__Pyx_PyLong_BoolEqObjC(__pyx_v_c, __pyx_mstate_global->__pyx_int_116, 0x74, 0)); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 98, __pyx_L1_error) if (__pyx_t_4) { } else { __pyx_t_1 = __pyx_t_4; goto __pyx_L40_bool_binop_done; } __pyx_t_4 = ((__pyx_cur_scope->__pyx_v_i + 4) <= __pyx_cur_scope->__pyx_v_n); if (__pyx_t_4) { } else { __pyx_t_1 = __pyx_t_4; goto __pyx_L40_bool_binop_done; } if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 98, __pyx_L1_error) } if (unlikely(__pyx_cur_scope->__pyx_v_b == Py_None)) { PyErr_SetString(PyExc_TypeError, "'NoneType' object is not subscriptable"); __PYX_ERR(0, 98, __pyx_L1_error) } __pyx_t_5 = PySequence_GetSlice(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, (__pyx_cur_scope->__pyx_v_i + 4)); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 98, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_5); __pyx_t_4 = (__Pyx_PyBytes_Equals(__pyx_t_5, __pyx_mstate_global->__pyx_n_b_true, Py_EQ)); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 98, __pyx_L1_error) __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; __pyx_t_1 = __pyx_t_4; __pyx_L40_bool_binop_done:; if (__pyx_t_1) { /* … */ }
+099: i += 4
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 4);
+100: return True
__Pyx_XDECREF(__pyx_r); __Pyx_INCREF(Py_True); __pyx_r = Py_True; goto __pyx_L0;
+101: if c == 102 and i + 5 <= n and b[i:i+5] == b'false':
__pyx_t_4 = (__Pyx_PyLong_BoolEqObjC(__pyx_v_c, __pyx_mstate_global->__pyx_int_102, 0x66, 0)); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 101, __pyx_L1_error) if (__pyx_t_4) { } else { __pyx_t_1 = __pyx_t_4; goto __pyx_L44_bool_binop_done; } __pyx_t_4 = ((__pyx_cur_scope->__pyx_v_i + 5) <= __pyx_cur_scope->__pyx_v_n); if (__pyx_t_4) { } else { __pyx_t_1 = __pyx_t_4; goto __pyx_L44_bool_binop_done; } if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 101, __pyx_L1_error) } if (unlikely(__pyx_cur_scope->__pyx_v_b == Py_None)) { PyErr_SetString(PyExc_TypeError, "'NoneType' object is not subscriptable"); __PYX_ERR(0, 101, __pyx_L1_error) } __pyx_t_5 = PySequence_GetSlice(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, (__pyx_cur_scope->__pyx_v_i + 5)); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 101, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_5); __pyx_t_4 = (__Pyx_PyBytes_Equals(__pyx_t_5, __pyx_mstate_global->__pyx_n_b_false, Py_EQ)); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 101, __pyx_L1_error) __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; __pyx_t_1 = __pyx_t_4; __pyx_L44_bool_binop_done:; if (__pyx_t_1) { /* … */ }
+102: i += 5
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 5);
+103: return False
__Pyx_XDECREF(__pyx_r); __Pyx_INCREF(Py_False); __pyx_r = Py_False; goto __pyx_L0;
104:
105: # number
+106: if c == 45 or (48 <= c <= 57):
__pyx_t_4 = (__Pyx_PyLong_BoolEqObjC(__pyx_v_c, __pyx_mstate_global->__pyx_int_45, 45, 0)); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 106, __pyx_L1_error) if (!__pyx_t_4) { } else { __pyx_t_1 = __pyx_t_4; goto __pyx_L48_bool_binop_done; } __pyx_t_5 = PyObject_RichCompare(__pyx_mstate_global->__pyx_int_48, __pyx_v_c, Py_LE); __Pyx_XGOTREF(__pyx_t_5); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 106, __pyx_L1_error) if (__Pyx_PyObject_IsTrue(__pyx_t_5)) { __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = PyObject_RichCompare(__pyx_v_c, __pyx_mstate_global->__pyx_int_57, Py_LE); __Pyx_XGOTREF(__pyx_t_5); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 106, __pyx_L1_error) } __pyx_t_4 = __Pyx_PyObject_IsTrue(__pyx_t_5); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 106, __pyx_L1_error) __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; __pyx_t_1 = __pyx_t_4; __pyx_L48_bool_binop_done:; if (__pyx_t_1) { /* … */ }
+107: start = i
__pyx_v_start = __pyx_cur_scope->__pyx_v_i;
+108: if c == 45:
__pyx_t_1 = (__Pyx_PyLong_BoolEqObjC(__pyx_v_c, __pyx_mstate_global->__pyx_int_45, 45, 0)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 108, __pyx_L1_error) if (__pyx_t_1) { /* … */ }
+109: i += 1
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 1);
+110: while i < n and 48 <= b[i] <= 57:
while (1) {
__pyx_t_4 = (__pyx_cur_scope->__pyx_v_i < __pyx_cur_scope->__pyx_v_n);
if (__pyx_t_4) {
} else {
__pyx_t_1 = __pyx_t_4;
goto __pyx_L53_bool_binop_done;
}
if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 110, __pyx_L1_error) }
__pyx_t_5 = __Pyx_GetItemInt(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, Py_ssize_t, 1, PyLong_FromSsize_t, 0, 0, 0, 1); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 110, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_5);
__pyx_t_6 = PyObject_RichCompare(__pyx_mstate_global->__pyx_int_48, __pyx_t_5, Py_LE); __Pyx_XGOTREF(__pyx_t_6); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 110, __pyx_L1_error)
if (__Pyx_PyObject_IsTrue(__pyx_t_6)) {
__Pyx_DECREF(__pyx_t_6);
__pyx_t_6 = PyObject_RichCompare(__pyx_t_5, __pyx_mstate_global->__pyx_int_57, Py_LE); __Pyx_XGOTREF(__pyx_t_6); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 110, __pyx_L1_error)
}
__Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
__pyx_t_4 = __Pyx_PyObject_IsTrue(__pyx_t_6); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 110, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
__pyx_t_1 = __pyx_t_4;
__pyx_L53_bool_binop_done:;
if (!__pyx_t_1) break;
+111: i += 1
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 1);
}
+112: if i < n and b[i] == 46:
__pyx_t_4 = (__pyx_cur_scope->__pyx_v_i < __pyx_cur_scope->__pyx_v_n);
if (__pyx_t_4) {
} else {
__pyx_t_1 = __pyx_t_4;
goto __pyx_L56_bool_binop_done;
}
if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 112, __pyx_L1_error) }
__pyx_t_6 = __Pyx_GetItemInt(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, Py_ssize_t, 1, PyLong_FromSsize_t, 0, 0, 0, 1); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 112, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_6);
__pyx_t_4 = (__Pyx_PyLong_BoolEqObjC(__pyx_t_6, __pyx_mstate_global->__pyx_int_46, 46, 0)); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 112, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
__pyx_t_1 = __pyx_t_4;
__pyx_L56_bool_binop_done:;
if (__pyx_t_1) {
/* … */
}
+113: i += 1
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 1);
+114: while i < n and 48 <= b[i] <= 57:
while (1) {
__pyx_t_4 = (__pyx_cur_scope->__pyx_v_i < __pyx_cur_scope->__pyx_v_n);
if (__pyx_t_4) {
} else {
__pyx_t_1 = __pyx_t_4;
goto __pyx_L60_bool_binop_done;
}
if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 114, __pyx_L1_error) }
__pyx_t_6 = __Pyx_GetItemInt(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, Py_ssize_t, 1, PyLong_FromSsize_t, 0, 0, 0, 1); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 114, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_6);
__pyx_t_5 = PyObject_RichCompare(__pyx_mstate_global->__pyx_int_48, __pyx_t_6, Py_LE); __Pyx_XGOTREF(__pyx_t_5); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 114, __pyx_L1_error)
if (__Pyx_PyObject_IsTrue(__pyx_t_5)) {
__Pyx_DECREF(__pyx_t_5);
__pyx_t_5 = PyObject_RichCompare(__pyx_t_6, __pyx_mstate_global->__pyx_int_57, Py_LE); __Pyx_XGOTREF(__pyx_t_5); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 114, __pyx_L1_error)
}
__Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
__pyx_t_4 = __Pyx_PyObject_IsTrue(__pyx_t_5); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 114, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
__pyx_t_1 = __pyx_t_4;
__pyx_L60_bool_binop_done:;
if (!__pyx_t_1) break;
+115: i += 1
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 1);
}
+116: if i < n and (b[i] == 101 or b[i] == 69):
__pyx_t_4 = (__pyx_cur_scope->__pyx_v_i < __pyx_cur_scope->__pyx_v_n);
if (__pyx_t_4) {
} else {
__pyx_t_1 = __pyx_t_4;
goto __pyx_L63_bool_binop_done;
}
if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 116, __pyx_L1_error) }
__pyx_t_5 = __Pyx_GetItemInt(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, Py_ssize_t, 1, PyLong_FromSsize_t, 0, 0, 0, 1); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 116, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_5);
__pyx_t_4 = (__Pyx_PyLong_BoolEqObjC(__pyx_t_5, __pyx_mstate_global->__pyx_int_101, 0x65, 0)); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 116, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
if (!__pyx_t_4) {
} else {
__pyx_t_1 = __pyx_t_4;
goto __pyx_L63_bool_binop_done;
}
if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 116, __pyx_L1_error) }
__pyx_t_5 = __Pyx_GetItemInt(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, Py_ssize_t, 1, PyLong_FromSsize_t, 0, 0, 0, 1); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 116, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_5);
__pyx_t_4 = (__Pyx_PyLong_BoolEqObjC(__pyx_t_5, __pyx_mstate_global->__pyx_int_69, 69, 0)); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 116, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
__pyx_t_1 = __pyx_t_4;
__pyx_L63_bool_binop_done:;
if (__pyx_t_1) {
/* … */
}
+117: i += 1
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 1);
+118: if i < n and (b[i] == 43 or b[i] == 45):
__pyx_t_4 = (__pyx_cur_scope->__pyx_v_i < __pyx_cur_scope->__pyx_v_n);
if (__pyx_t_4) {
} else {
__pyx_t_1 = __pyx_t_4;
goto __pyx_L67_bool_binop_done;
}
if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 118, __pyx_L1_error) }
__pyx_t_5 = __Pyx_GetItemInt(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, Py_ssize_t, 1, PyLong_FromSsize_t, 0, 0, 0, 1); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 118, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_5);
__pyx_t_4 = (__Pyx_PyLong_BoolEqObjC(__pyx_t_5, __pyx_mstate_global->__pyx_int_43, 43, 0)); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 118, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
if (!__pyx_t_4) {
} else {
__pyx_t_1 = __pyx_t_4;
goto __pyx_L67_bool_binop_done;
}
if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 118, __pyx_L1_error) }
__pyx_t_5 = __Pyx_GetItemInt(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, Py_ssize_t, 1, PyLong_FromSsize_t, 0, 0, 0, 1); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 118, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_5);
__pyx_t_4 = (__Pyx_PyLong_BoolEqObjC(__pyx_t_5, __pyx_mstate_global->__pyx_int_45, 45, 0)); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 118, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
__pyx_t_1 = __pyx_t_4;
__pyx_L67_bool_binop_done:;
if (__pyx_t_1) {
/* … */
}
+119: i += 1
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 1);
+120: while i < n and 48 <= b[i] <= 57:
while (1) {
__pyx_t_4 = (__pyx_cur_scope->__pyx_v_i < __pyx_cur_scope->__pyx_v_n);
if (__pyx_t_4) {
} else {
__pyx_t_1 = __pyx_t_4;
goto __pyx_L72_bool_binop_done;
}
if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 120, __pyx_L1_error) }
__pyx_t_5 = __Pyx_GetItemInt(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, Py_ssize_t, 1, PyLong_FromSsize_t, 0, 0, 0, 1); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 120, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_5);
__pyx_t_6 = PyObject_RichCompare(__pyx_mstate_global->__pyx_int_48, __pyx_t_5, Py_LE); __Pyx_XGOTREF(__pyx_t_6); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 120, __pyx_L1_error)
if (__Pyx_PyObject_IsTrue(__pyx_t_6)) {
__Pyx_DECREF(__pyx_t_6);
__pyx_t_6 = PyObject_RichCompare(__pyx_t_5, __pyx_mstate_global->__pyx_int_57, Py_LE); __Pyx_XGOTREF(__pyx_t_6); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 120, __pyx_L1_error)
}
__Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
__pyx_t_4 = __Pyx_PyObject_IsTrue(__pyx_t_6); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 120, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
__pyx_t_1 = __pyx_t_4;
__pyx_L72_bool_binop_done:;
if (!__pyx_t_1) break;
+121: i += 1
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 1);
}
+122: return float(b[start:i].decode('ascii'))
__Pyx_XDECREF(__pyx_r); if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 122, __pyx_L1_error) } if (unlikely(__pyx_cur_scope->__pyx_v_b == Py_None)) { PyErr_SetString(PyExc_TypeError, "'NoneType' object is not subscriptable"); __PYX_ERR(0, 122, __pyx_L1_error) } __pyx_t_6 = __Pyx_decode_bytes(__pyx_cur_scope->__pyx_v_b, __pyx_v_start, __pyx_cur_scope->__pyx_v_i, NULL, NULL, PyUnicode_DecodeASCII); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 122, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_6); __pyx_t_14 = __Pyx_PyUnicode_AsDouble(__pyx_t_6); if (unlikely(__pyx_t_14 == ((double)((double)-1)) && PyErr_Occurred())) __PYX_ERR(0, 122, __pyx_L1_error) __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0; __pyx_t_6 = PyFloat_FromDouble(__pyx_t_14); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 122, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_6); __pyx_r = __pyx_t_6; __pyx_t_6 = 0; goto __pyx_L0;
+123: s = b[start:i].decode('ascii')
if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 123, __pyx_L1_error) }
if (unlikely(__pyx_cur_scope->__pyx_v_b == Py_None)) {
PyErr_SetString(PyExc_TypeError, "'NoneType' object is not subscriptable");
__PYX_ERR(0, 123, __pyx_L1_error)
}
__pyx_t_6 = __Pyx_decode_bytes(__pyx_cur_scope->__pyx_v_b, __pyx_v_start, __pyx_cur_scope->__pyx_v_i, NULL, NULL, PyUnicode_DecodeASCII); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 123, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_6);
__pyx_v_s = ((PyObject*)__pyx_t_6);
__pyx_t_6 = 0;
+124: if '.' in s or 'e' in s or 'E' in s:
__pyx_t_4 = (__Pyx_PyUnicode_ContainsTF(__pyx_mstate_global->__pyx_kp_u__10, __pyx_v_s, Py_EQ)); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 124, __pyx_L1_error) if (!__pyx_t_4) { } else { __pyx_t_1 = __pyx_t_4; goto __pyx_L75_bool_binop_done; } __pyx_t_4 = (__Pyx_PyUnicode_ContainsTF(__pyx_mstate_global->__pyx_n_u_e, __pyx_v_s, Py_EQ)); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 124, __pyx_L1_error) if (!__pyx_t_4) { } else { __pyx_t_1 = __pyx_t_4; goto __pyx_L75_bool_binop_done; } __pyx_t_4 = (__Pyx_PyUnicode_ContainsTF(__pyx_mstate_global->__pyx_n_u_E, __pyx_v_s, Py_EQ)); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 124, __pyx_L1_error) __pyx_t_1 = __pyx_t_4; __pyx_L75_bool_binop_done:; if (__pyx_t_1) { /* … */ }
+125: return float(s)
__Pyx_XDECREF(__pyx_r); __pyx_t_14 = __Pyx_PyUnicode_AsDouble(__pyx_v_s); if (unlikely(__pyx_t_14 == ((double)((double)-1)) && PyErr_Occurred())) __PYX_ERR(0, 125, __pyx_L1_error) __pyx_t_6 = PyFloat_FromDouble(__pyx_t_14); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 125, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_6); __pyx_r = __pyx_t_6; __pyx_t_6 = 0; goto __pyx_L0;
126: else:
+127: try:
/*else*/ {
{
/*try:*/ {
/* … */
}
__pyx_L78_error:;
__Pyx_XDECREF(__pyx_t_13); __pyx_t_13 = 0;
__Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;
__Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
__Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
/* … */
__pyx_L80_except_error:;
__Pyx_XGIVEREF(__pyx_t_11);
__Pyx_XGIVEREF(__pyx_t_10);
__Pyx_XGIVEREF(__pyx_t_9);
__Pyx_ExceptionReset(__pyx_t_11, __pyx_t_10, __pyx_t_9);
goto __pyx_L1_error;
__pyx_L82_try_return:;
__Pyx_XGIVEREF(__pyx_t_11);
__Pyx_XGIVEREF(__pyx_t_10);
__Pyx_XGIVEREF(__pyx_t_9);
__Pyx_ExceptionReset(__pyx_t_11, __pyx_t_10, __pyx_t_9);
goto __pyx_L0;
__pyx_L81_except_return:;
__Pyx_XGIVEREF(__pyx_t_11);
__Pyx_XGIVEREF(__pyx_t_10);
__Pyx_XGIVEREF(__pyx_t_9);
__Pyx_ExceptionReset(__pyx_t_11, __pyx_t_10, __pyx_t_9);
goto __pyx_L0;
}
}
+128: return int(s)
__Pyx_XDECREF(__pyx_r); __pyx_t_6 = __Pyx_PyNumber_Int(__pyx_v_s); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 128, __pyx_L78_error) __Pyx_GOTREF(__pyx_t_6); __pyx_r = __pyx_t_6; __pyx_t_6 = 0; goto __pyx_L82_try_return;
+129: except Exception:
__pyx_t_12 = __Pyx_PyErr_ExceptionMatches(((PyObject *)(((PyTypeObject*)PyExc_Exception)))); if (__pyx_t_12) { __Pyx_AddTraceback("rugo.jsonl._parse_array_from_bytes.parse_value", __pyx_clineno, __pyx_lineno, __pyx_filename); if (__Pyx_GetException(&__pyx_t_6, &__pyx_t_5, &__pyx_t_3) < 0) __PYX_ERR(0, 129, __pyx_L80_except_error) __Pyx_XGOTREF(__pyx_t_6); __Pyx_XGOTREF(__pyx_t_5); __Pyx_XGOTREF(__pyx_t_3);
+130: return float(s)
__Pyx_XDECREF(__pyx_r); __pyx_t_14 = __Pyx_PyUnicode_AsDouble(__pyx_v_s); if (unlikely(__pyx_t_14 == ((double)((double)-1)) && PyErr_Occurred())) __PYX_ERR(0, 130, __pyx_L80_except_error) __pyx_t_13 = PyFloat_FromDouble(__pyx_t_14); if (unlikely(!__pyx_t_13)) __PYX_ERR(0, 130, __pyx_L80_except_error) __Pyx_GOTREF(__pyx_t_13); __pyx_r = __pyx_t_13; __pyx_t_13 = 0; __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0; goto __pyx_L81_except_return; } goto __pyx_L80_except_error;
131:
132: # array
+133: if c == 91: # '['
__pyx_t_1 = (__Pyx_PyLong_BoolEqObjC(__pyx_v_c, __pyx_mstate_global->__pyx_int_91, 91, 0)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 133, __pyx_L1_error) if (__pyx_t_1) { /* … */ }
134: # parse nested array
+135: i += 1
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 1);
+136: res = []
__pyx_t_3 = PyList_New(0); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 136, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_3); __pyx_v_res = ((PyObject*)__pyx_t_3); __pyx_t_3 = 0;
137: # skip whitespace
+138: while i < n and b[i] in (32,9,10,13): i += 1
while (1) {
__pyx_t_4 = (__pyx_cur_scope->__pyx_v_i < __pyx_cur_scope->__pyx_v_n);
if (__pyx_t_4) {
} else {
__pyx_t_1 = __pyx_t_4;
goto __pyx_L89_bool_binop_done;
}
if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 138, __pyx_L1_error) }
__pyx_t_3 = __Pyx_GetItemInt(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, Py_ssize_t, 1, PyLong_FromSsize_t, 0, 0, 0, 1); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 138, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_3);
__pyx_t_2 = (__Pyx_PyLong_BoolEqObjC(__pyx_t_3, __pyx_mstate_global->__pyx_int_32, 32, 0)); if (unlikely((__pyx_t_2 < 0))) __PYX_ERR(0, 138, __pyx_L1_error)
if (!__pyx_t_2) {
} else {
__pyx_t_4 = __pyx_t_2;
goto __pyx_L91_bool_binop_done;
}
__pyx_t_2 = (__Pyx_PyLong_BoolEqObjC(__pyx_t_3, __pyx_mstate_global->__pyx_int_9, 9, 0)); if (unlikely((__pyx_t_2 < 0))) __PYX_ERR(0, 138, __pyx_L1_error)
if (!__pyx_t_2) {
} else {
__pyx_t_4 = __pyx_t_2;
goto __pyx_L91_bool_binop_done;
}
__pyx_t_2 = (__Pyx_PyLong_BoolEqObjC(__pyx_t_3, __pyx_mstate_global->__pyx_int_10, 10, 0)); if (unlikely((__pyx_t_2 < 0))) __PYX_ERR(0, 138, __pyx_L1_error)
if (!__pyx_t_2) {
} else {
__pyx_t_4 = __pyx_t_2;
goto __pyx_L91_bool_binop_done;
}
__pyx_t_2 = (__Pyx_PyLong_BoolEqObjC(__pyx_t_3, __pyx_mstate_global->__pyx_int_13, 13, 0)); if (unlikely((__pyx_t_2 < 0))) __PYX_ERR(0, 138, __pyx_L1_error)
__pyx_t_4 = __pyx_t_2;
__pyx_L91_bool_binop_done:;
__Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
__pyx_t_2 = __pyx_t_4;
__pyx_t_1 = __pyx_t_2;
__pyx_L89_bool_binop_done:;
if (!__pyx_t_1) break;
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 1);
}
+139: if i < n and b[i] == 93:
__pyx_t_2 = (__pyx_cur_scope->__pyx_v_i < __pyx_cur_scope->__pyx_v_n);
if (__pyx_t_2) {
} else {
__pyx_t_1 = __pyx_t_2;
goto __pyx_L96_bool_binop_done;
}
if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 139, __pyx_L1_error) }
__pyx_t_3 = __Pyx_GetItemInt(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, Py_ssize_t, 1, PyLong_FromSsize_t, 0, 0, 0, 1); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 139, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_3);
__pyx_t_2 = (__Pyx_PyLong_BoolEqObjC(__pyx_t_3, __pyx_mstate_global->__pyx_int_93, 93, 0)); if (unlikely((__pyx_t_2 < 0))) __PYX_ERR(0, 139, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
__pyx_t_1 = __pyx_t_2;
__pyx_L96_bool_binop_done:;
if (__pyx_t_1) {
/* … */
}
+140: i += 1
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 1);
+141: return res
__Pyx_XDECREF(__pyx_r); __Pyx_INCREF(__pyx_v_res); __pyx_r = __pyx_v_res; goto __pyx_L0;
+142: while True:
while (1) {
+143: val = parse_value()
if (unlikely(!__pyx_cur_scope->__pyx_v_parse_value)) { __Pyx_RaiseClosureNameError("parse_value"); __PYX_ERR(0, 143, __pyx_L1_error) }
__pyx_t_3 = __pyx_pf_4rugo_5jsonl_23_parse_array_from_bytes_parse_value(__pyx_cur_scope->__pyx_v_parse_value); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 143, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_3);
__Pyx_XDECREF_SET(__pyx_v_val, __pyx_t_3);
__pyx_t_3 = 0;
+144: res.append(val)
__pyx_t_8 = __Pyx_PyList_Append(__pyx_v_res, __pyx_v_val); if (unlikely(__pyx_t_8 == ((int)-1))) __PYX_ERR(0, 144, __pyx_L1_error)
+145: while i < n and b[i] in (32,9,10,13): i += 1
while (1) {
__pyx_t_2 = (__pyx_cur_scope->__pyx_v_i < __pyx_cur_scope->__pyx_v_n);
if (__pyx_t_2) {
} else {
__pyx_t_1 = __pyx_t_2;
goto __pyx_L102_bool_binop_done;
}
if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 145, __pyx_L1_error) }
__pyx_t_3 = __Pyx_GetItemInt(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, Py_ssize_t, 1, PyLong_FromSsize_t, 0, 0, 0, 1); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 145, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_3);
__pyx_t_4 = (__Pyx_PyLong_BoolEqObjC(__pyx_t_3, __pyx_mstate_global->__pyx_int_32, 32, 0)); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 145, __pyx_L1_error)
if (!__pyx_t_4) {
} else {
__pyx_t_2 = __pyx_t_4;
goto __pyx_L104_bool_binop_done;
}
__pyx_t_4 = (__Pyx_PyLong_BoolEqObjC(__pyx_t_3, __pyx_mstate_global->__pyx_int_9, 9, 0)); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 145, __pyx_L1_error)
if (!__pyx_t_4) {
} else {
__pyx_t_2 = __pyx_t_4;
goto __pyx_L104_bool_binop_done;
}
__pyx_t_4 = (__Pyx_PyLong_BoolEqObjC(__pyx_t_3, __pyx_mstate_global->__pyx_int_10, 10, 0)); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 145, __pyx_L1_error)
if (!__pyx_t_4) {
} else {
__pyx_t_2 = __pyx_t_4;
goto __pyx_L104_bool_binop_done;
}
__pyx_t_4 = (__Pyx_PyLong_BoolEqObjC(__pyx_t_3, __pyx_mstate_global->__pyx_int_13, 13, 0)); if (unlikely((__pyx_t_4 < 0))) __PYX_ERR(0, 145, __pyx_L1_error)
__pyx_t_2 = __pyx_t_4;
__pyx_L104_bool_binop_done:;
__Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
__pyx_t_4 = __pyx_t_2;
__pyx_t_1 = __pyx_t_4;
__pyx_L102_bool_binop_done:;
if (!__pyx_t_1) break;
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 1);
}
+146: if i >= n:
__pyx_t_1 = (__pyx_cur_scope->__pyx_v_i >= __pyx_cur_scope->__pyx_v_n);
if (unlikely(__pyx_t_1)) {
/* … */
}
+147: raise ValueError('unterminated array')
__pyx_t_5 = NULL;
__Pyx_INCREF(__pyx_builtin_ValueError);
__pyx_t_6 = __pyx_builtin_ValueError;
__pyx_t_7 = 1;
{
PyObject *__pyx_callargs[2] = {__pyx_t_5, __pyx_mstate_global->__pyx_kp_u_unterminated_array};
__pyx_t_3 = __Pyx_PyObject_FastCall(__pyx_t_6, __pyx_callargs+__pyx_t_7, (2-__pyx_t_7) | (__pyx_t_7*__Pyx_PY_VECTORCALL_ARGUMENTS_OFFSET));
__Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
__Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 147, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_3);
}
__Pyx_Raise(__pyx_t_3, 0, 0, 0);
__Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
__PYX_ERR(0, 147, __pyx_L1_error)
+148: if b[i] == 44:
if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 148, __pyx_L1_error) }
__pyx_t_3 = __Pyx_GetItemInt(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, Py_ssize_t, 1, PyLong_FromSsize_t, 0, 0, 0, 1); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 148, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_3);
__pyx_t_1 = (__Pyx_PyLong_BoolEqObjC(__pyx_t_3, __pyx_mstate_global->__pyx_int_44, 44, 0)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 148, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
if (__pyx_t_1) {
/* … */
}
+149: i += 1
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 1);
+150: continue
goto __pyx_L98_continue;
+151: elif b[i] == 93:
if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 151, __pyx_L1_error) }
__pyx_t_3 = __Pyx_GetItemInt(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, Py_ssize_t, 1, PyLong_FromSsize_t, 0, 0, 0, 1); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 151, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_3);
__pyx_t_1 = (__Pyx_PyLong_BoolEqObjC(__pyx_t_3, __pyx_mstate_global->__pyx_int_93, 93, 0)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 151, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
if (likely(__pyx_t_1)) {
/* … */
}
+152: i += 1
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 1);
+153: break
goto __pyx_L99_break;
154: else:
+155: raise ValueError('invalid array separator')
/*else*/ {
__pyx_t_6 = NULL;
__Pyx_INCREF(__pyx_builtin_ValueError);
__pyx_t_5 = __pyx_builtin_ValueError;
__pyx_t_7 = 1;
{
PyObject *__pyx_callargs[2] = {__pyx_t_6, __pyx_mstate_global->__pyx_kp_u_invalid_array_separator};
__pyx_t_3 = __Pyx_PyObject_FastCall(__pyx_t_5, __pyx_callargs+__pyx_t_7, (2-__pyx_t_7) | (__pyx_t_7*__Pyx_PY_VECTORCALL_ARGUMENTS_OFFSET));
__Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
__Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 155, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_3);
}
__Pyx_Raise(__pyx_t_3, 0, 0, 0);
__Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
__PYX_ERR(0, 155, __pyx_L1_error)
}
__pyx_L98_continue:;
}
__pyx_L99_break:;
+156: return res
__Pyx_XDECREF(__pyx_r); __Pyx_INCREF(__pyx_v_res); __pyx_r = __pyx_v_res; goto __pyx_L0;
157:
158: # object: return raw bytes slice for object
+159: if c == 123: # '{'
__pyx_t_1 = (__Pyx_PyLong_BoolEqObjC(__pyx_v_c, __pyx_mstate_global->__pyx_int_123, 0x7B, 0)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 159, __pyx_L1_error) if (__pyx_t_1) { /* … */ }
+160: start = i
__pyx_v_start = __pyx_cur_scope->__pyx_v_i;
+161: depth = 0
__Pyx_INCREF(__pyx_mstate_global->__pyx_int_0);
__pyx_v_depth = __pyx_mstate_global->__pyx_int_0;
+162: while i < n:
while (1) {
__pyx_t_1 = (__pyx_cur_scope->__pyx_v_i < __pyx_cur_scope->__pyx_v_n);
if (!__pyx_t_1) break;
+163: ch = b[i]
if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 163, __pyx_L1_error) }
__pyx_t_3 = __Pyx_GetItemInt(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, Py_ssize_t, 1, PyLong_FromSsize_t, 0, 0, 0, 1); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 163, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_3);
__Pyx_XDECREF_SET(__pyx_v_ch, __pyx_t_3);
__pyx_t_3 = 0;
+164: if ch == 34:
__pyx_t_1 = (__Pyx_PyLong_BoolEqObjC(__pyx_v_ch, __pyx_mstate_global->__pyx_int_34, 34, 0)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 164, __pyx_L1_error) if (__pyx_t_1) { /* … */ }
165: # skip string
+166: i += 1
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 1);
+167: while i < n:
while (1) {
__pyx_t_1 = (__pyx_cur_scope->__pyx_v_i < __pyx_cur_scope->__pyx_v_n);
if (!__pyx_t_1) break;
+168: if b[i] == 92:
if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 168, __pyx_L1_error) }
__pyx_t_3 = __Pyx_GetItemInt(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, Py_ssize_t, 1, PyLong_FromSsize_t, 0, 0, 0, 1); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 168, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_3);
__pyx_t_1 = (__Pyx_PyLong_BoolEqObjC(__pyx_t_3, __pyx_mstate_global->__pyx_int_92, 92, 0)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 168, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
if (__pyx_t_1) {
/* … */
goto __pyx_L116;
}
+169: i += 2
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 2);
+170: elif b[i] == 34:
if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 170, __pyx_L1_error) }
__pyx_t_3 = __Pyx_GetItemInt(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, Py_ssize_t, 1, PyLong_FromSsize_t, 0, 0, 0, 1); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 170, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_3);
__pyx_t_1 = (__Pyx_PyLong_BoolEqObjC(__pyx_t_3, __pyx_mstate_global->__pyx_int_34, 34, 0)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 170, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
if (__pyx_t_1) {
/* … */
}
+171: i += 1
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 1);
+172: break
goto __pyx_L115_break;
173: else:
+174: i += 1
/*else*/ {
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 1);
}
__pyx_L116:;
}
__pyx_L115_break:;
+175: continue
goto __pyx_L111_continue;
+176: if ch == 123:
__pyx_t_1 = (__Pyx_PyLong_BoolEqObjC(__pyx_v_ch, __pyx_mstate_global->__pyx_int_123, 0x7B, 0)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 176, __pyx_L1_error) if (__pyx_t_1) { /* … */ goto __pyx_L117; }
+177: depth += 1
__pyx_t_3 = __Pyx_PyLong_AddObjC(__pyx_v_depth, __pyx_mstate_global->__pyx_int_1, 1, 1, 0); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 177, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_3); __Pyx_DECREF_SET(__pyx_v_depth, __pyx_t_3); __pyx_t_3 = 0;
+178: elif ch == 125:
__pyx_t_1 = (__Pyx_PyLong_BoolEqObjC(__pyx_v_ch, __pyx_mstate_global->__pyx_int_125, 0x7D, 0)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 178, __pyx_L1_error) if (__pyx_t_1) { /* … */ } __pyx_L117:;
+179: depth -= 1
__pyx_t_3 = __Pyx_PyLong_SubtractObjC(__pyx_v_depth, __pyx_mstate_global->__pyx_int_1, 1, 1, 0); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 179, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_3); __Pyx_DECREF_SET(__pyx_v_depth, __pyx_t_3); __pyx_t_3 = 0;
+180: if depth == 0:
__pyx_t_1 = (__Pyx_PyLong_BoolEqObjC(__pyx_v_depth, __pyx_mstate_global->__pyx_int_0, 0, 0)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 180, __pyx_L1_error) if (__pyx_t_1) { /* … */ }
+181: i += 1
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 1);
+182: return b[start:i]
__Pyx_XDECREF(__pyx_r); if (unlikely(!__pyx_cur_scope->__pyx_v_b)) { __Pyx_RaiseClosureNameError("b"); __PYX_ERR(0, 182, __pyx_L1_error) } if (unlikely(__pyx_cur_scope->__pyx_v_b == Py_None)) { PyErr_SetString(PyExc_TypeError, "'NoneType' object is not subscriptable"); __PYX_ERR(0, 182, __pyx_L1_error) } __pyx_t_3 = PySequence_GetSlice(__pyx_cur_scope->__pyx_v_b, __pyx_v_start, __pyx_cur_scope->__pyx_v_i); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 182, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_3); __pyx_r = __pyx_t_3; __pyx_t_3 = 0; goto __pyx_L0;
+183: i += 1
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 1);
__pyx_L111_continue:;
}
+184: raise ValueError('unterminated object')
__pyx_t_5 = NULL;
__Pyx_INCREF(__pyx_builtin_ValueError);
__pyx_t_6 = __pyx_builtin_ValueError;
__pyx_t_7 = 1;
{
PyObject *__pyx_callargs[2] = {__pyx_t_5, __pyx_mstate_global->__pyx_kp_u_unterminated_object};
__pyx_t_3 = __Pyx_PyObject_FastCall(__pyx_t_6, __pyx_callargs+__pyx_t_7, (2-__pyx_t_7) | (__pyx_t_7*__Pyx_PY_VECTORCALL_ARGUMENTS_OFFSET));
__Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
__Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 184, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_3);
}
__Pyx_Raise(__pyx_t_3, 0, 0, 0);
__Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
__PYX_ERR(0, 184, __pyx_L1_error)
185:
+186: raise ValueError('unexpected token at %d' % i)
__pyx_t_6 = NULL; __Pyx_INCREF(__pyx_builtin_ValueError); __pyx_t_5 = __pyx_builtin_ValueError; __pyx_t_13 = PyLong_FromSsize_t(__pyx_cur_scope->__pyx_v_i); if (unlikely(!__pyx_t_13)) __PYX_ERR(0, 186, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_13); __pyx_t_15 = PyUnicode_Format(__pyx_mstate_global->__pyx_kp_u_unexpected_token_at_d, __pyx_t_13); if (unlikely(!__pyx_t_15)) __PYX_ERR(0, 186, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_15); __Pyx_DECREF(__pyx_t_13); __pyx_t_13 = 0; __pyx_t_7 = 1; { PyObject *__pyx_callargs[2] = {__pyx_t_6, __pyx_t_15}; __pyx_t_3 = __Pyx_PyObject_FastCall(__pyx_t_5, __pyx_callargs+__pyx_t_7, (2-__pyx_t_7) | (__pyx_t_7*__Pyx_PY_VECTORCALL_ARGUMENTS_OFFSET)); __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0; __Pyx_DECREF(__pyx_t_15); __pyx_t_15 = 0; __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0; if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 186, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_3); } __Pyx_Raise(__pyx_t_3, 0, 0, 0); __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0; __PYX_ERR(0, 186, __pyx_L1_error)
187:
188: # top-level: expect '['
+189: while i < n and b[i] in (32,9,10,13): i += 1
while (1) {
__pyx_t_4 = (__pyx_cur_scope->__pyx_v_i < __pyx_cur_scope->__pyx_v_n);
if (__pyx_t_4) {
} else {
__pyx_t_3 = __pyx_t_4;
goto __pyx_L5_bool_binop_done;
}
__pyx_t_1 = __Pyx_GetItemInt(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, Py_ssize_t, 1, PyLong_FromSsize_t, 0, 0, 0, 1); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 189, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_1);
__pyx_t_5 = (__Pyx_PyLong_BoolEqObjC(__pyx_t_1, __pyx_mstate_global->__pyx_int_32, 32, 0)); if (unlikely((__pyx_t_5 < 0))) __PYX_ERR(0, 189, __pyx_L1_error)
if (!__pyx_t_5) {
} else {
__pyx_t_4 = __pyx_t_5;
goto __pyx_L7_bool_binop_done;
}
__pyx_t_5 = (__Pyx_PyLong_BoolEqObjC(__pyx_t_1, __pyx_mstate_global->__pyx_int_9, 9, 0)); if (unlikely((__pyx_t_5 < 0))) __PYX_ERR(0, 189, __pyx_L1_error)
if (!__pyx_t_5) {
} else {
__pyx_t_4 = __pyx_t_5;
goto __pyx_L7_bool_binop_done;
}
__pyx_t_5 = (__Pyx_PyLong_BoolEqObjC(__pyx_t_1, __pyx_mstate_global->__pyx_int_10, 10, 0)); if (unlikely((__pyx_t_5 < 0))) __PYX_ERR(0, 189, __pyx_L1_error)
if (!__pyx_t_5) {
} else {
__pyx_t_4 = __pyx_t_5;
goto __pyx_L7_bool_binop_done;
}
__pyx_t_5 = (__Pyx_PyLong_BoolEqObjC(__pyx_t_1, __pyx_mstate_global->__pyx_int_13, 13, 0)); if (unlikely((__pyx_t_5 < 0))) __PYX_ERR(0, 189, __pyx_L1_error)
__pyx_t_4 = __pyx_t_5;
__pyx_L7_bool_binop_done:;
__Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;
__pyx_t_5 = __pyx_t_4;
__pyx_t_3 = __pyx_t_5;
__pyx_L5_bool_binop_done:;
if (!__pyx_t_3) break;
__pyx_cur_scope->__pyx_v_i = (__pyx_cur_scope->__pyx_v_i + 1);
}
+190: if i >= n or b[i] != 91:
__pyx_t_5 = (__pyx_cur_scope->__pyx_v_i >= __pyx_cur_scope->__pyx_v_n);
if (!__pyx_t_5) {
} else {
__pyx_t_3 = __pyx_t_5;
goto __pyx_L12_bool_binop_done;
}
__pyx_t_1 = __Pyx_GetItemInt(__pyx_cur_scope->__pyx_v_b, __pyx_cur_scope->__pyx_v_i, Py_ssize_t, 1, PyLong_FromSsize_t, 0, 0, 0, 1); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 190, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_1);
__pyx_t_5 = (__Pyx_PyLong_BoolNeObjC(__pyx_t_1, __pyx_mstate_global->__pyx_int_91, 91, 0)); if (unlikely((__pyx_t_5 < 0))) __PYX_ERR(0, 190, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;
__pyx_t_3 = __pyx_t_5;
__pyx_L12_bool_binop_done:;
if (unlikely(__pyx_t_3)) {
/* … */
}
+191: raise ValueError('not an array')
__pyx_t_6 = NULL;
__Pyx_INCREF(__pyx_builtin_ValueError);
__pyx_t_7 = __pyx_builtin_ValueError;
__pyx_t_8 = 1;
{
PyObject *__pyx_callargs[2] = {__pyx_t_6, __pyx_mstate_global->__pyx_kp_u_not_an_array};
__pyx_t_1 = __Pyx_PyObject_FastCall(__pyx_t_7, __pyx_callargs+__pyx_t_8, (2-__pyx_t_8) | (__pyx_t_8*__Pyx_PY_VECTORCALL_ARGUMENTS_OFFSET));
__Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
__Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 191, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_1);
}
__Pyx_Raise(__pyx_t_1, 0, 0, 0);
__Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;
__PYX_ERR(0, 191, __pyx_L1_error)
+192: return parse_value()
__Pyx_XDECREF(__pyx_r); __pyx_t_1 = __pyx_pf_4rugo_5jsonl_23_parse_array_from_bytes_parse_value(__pyx_cur_scope->__pyx_v_parse_value); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 192, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_1); __pyx_r = __pyx_t_1; __pyx_t_1 = 0; goto __pyx_L0;
193:
194: cdef extern from "decode.hpp":
195: cdef enum JsonType:
196: pass
197: cdef cppclass ColumnSchema:
198: string name
199: JsonType type
200: bint nullable
201: JsonType element_type
202: cdef cppclass JsonlColumn:
203: vector[int64_t] int_values
204: vector[double] double_values
205: vector[string] string_values
206: vector[uint8_t] boolean_values
207: vector[uint8_t] null_mask
208: string type
209: bint success
210: cdef cppclass JsonlTable:
211: vector[JsonlColumn] columns
212: vector[string] column_names
213: size_t num_rows
214: bint success
215: vector[ColumnSchema] GetJsonlSchema(const uint8_t* data, size_t size, size_t sample_size) except +
216: JsonlTable ReadJsonl(const uint8_t* data, size_t size, const vector[string]& column_names) except +
217: JsonlTable ReadJsonl(const uint8_t* data, size_t size) except +
218: PyObject* ParseJsonSliceToPyObject(const uint8_t* data, size_t len, bint parse_objects)
219:
+220: def get_jsonl_schema(data, sample_size=25):
/* Python wrapper */ static PyObject *__pyx_pw_4rugo_5jsonl_3get_jsonl_schema(PyObject *__pyx_self, #if CYTHON_METH_FASTCALL PyObject *const *__pyx_args, Py_ssize_t __pyx_nargs, PyObject *__pyx_kwds #else PyObject *__pyx_args, PyObject *__pyx_kwds #endif ); /*proto*/ PyDoc_STRVAR(__pyx_doc_4rugo_5jsonl_2get_jsonl_schema, "\n Infer the schema of a JSONL dataset from a sample of the data.\n\n Parameters\n ----------\n data : bytes or object supporting the buffer protocol\n The JSONL data to analyze.\n sample_size : int, optional\n The number of rows to sample for schema inference (default: 25).\n\n Returns\n -------\n list of dict\n A list of dictionaries, each describing a column with keys:\n - 'name': str, the column name\n - 'type': str, the inferred type ('null', 'boolean', 'int64', 'double', 'bytes')\n - 'nullable': bool, whether the column can contain null values\n "); static PyMethodDef __pyx_mdef_4rugo_5jsonl_3get_jsonl_schema = {"get_jsonl_schema", (PyCFunction)(void(*)(void))(__Pyx_PyCFunction_FastCallWithKeywords)__pyx_pw_4rugo_5jsonl_3get_jsonl_schema, __Pyx_METH_FASTCALL|METH_KEYWORDS, __pyx_doc_4rugo_5jsonl_2get_jsonl_schema}; static PyObject *__pyx_pw_4rugo_5jsonl_3get_jsonl_schema(PyObject *__pyx_self, #if CYTHON_METH_FASTCALL PyObject *const *__pyx_args, Py_ssize_t __pyx_nargs, PyObject *__pyx_kwds #else PyObject *__pyx_args, PyObject *__pyx_kwds #endif ) { PyObject *__pyx_v_data = 0; PyObject *__pyx_v_sample_size = 0; #if !CYTHON_METH_FASTCALL CYTHON_UNUSED Py_ssize_t __pyx_nargs; #endif CYTHON_UNUSED PyObject *const *__pyx_kwvalues; PyObject *__pyx_r = 0; __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("get_jsonl_schema (wrapper)", 0); #if !CYTHON_METH_FASTCALL #if CYTHON_ASSUME_SAFE_SIZE __pyx_nargs = PyTuple_GET_SIZE(__pyx_args); #else __pyx_nargs = PyTuple_Size(__pyx_args); if (unlikely(__pyx_nargs < 0)) return NULL; #endif #endif __pyx_kwvalues = __Pyx_KwValues_FASTCALL(__pyx_args, __pyx_nargs); { PyObject ** const __pyx_pyargnames[] = {&__pyx_mstate_global->__pyx_n_u_data,&__pyx_mstate_global->__pyx_n_u_sample_size,0}; PyObject* values[2] = {0,0}; const Py_ssize_t __pyx_kwds_len = (__pyx_kwds) ? __Pyx_NumKwargs_FASTCALL(__pyx_kwds) : 0; if (unlikely(__pyx_kwds_len) < 0) __PYX_ERR(0, 220, __pyx_L3_error) if (__pyx_kwds_len > 0) { switch (__pyx_nargs) { case 2: values[1] = __Pyx_ArgRef_FASTCALL(__pyx_args, 1); if (!CYTHON_ASSUME_SAFE_MACROS && unlikely(!values[1])) __PYX_ERR(0, 220, __pyx_L3_error) CYTHON_FALLTHROUGH; case 1: values[0] = __Pyx_ArgRef_FASTCALL(__pyx_args, 0); if (!CYTHON_ASSUME_SAFE_MACROS && unlikely(!values[0])) __PYX_ERR(0, 220, __pyx_L3_error) CYTHON_FALLTHROUGH; case 0: break; default: goto __pyx_L5_argtuple_error; } const Py_ssize_t kwd_pos_args = __pyx_nargs; if (__Pyx_ParseKeywords(__pyx_kwds, __pyx_kwvalues, __pyx_pyargnames, 0, values, kwd_pos_args, __pyx_kwds_len, "get_jsonl_schema", 0) < (0)) __PYX_ERR(0, 220, __pyx_L3_error) if (!values[1]) values[1] = __Pyx_NewRef(((PyObject *)((PyObject*)__pyx_mstate_global->__pyx_int_25))); for (Py_ssize_t i = __pyx_nargs; i < 1; i++) { if (unlikely(!values[i])) { __Pyx_RaiseArgtupleInvalid("get_jsonl_schema", 0, 1, 2, i); __PYX_ERR(0, 220, __pyx_L3_error) } } } else { switch (__pyx_nargs) { case 2: values[1] = __Pyx_ArgRef_FASTCALL(__pyx_args, 1); if (!CYTHON_ASSUME_SAFE_MACROS && unlikely(!values[1])) __PYX_ERR(0, 220, __pyx_L3_error) CYTHON_FALLTHROUGH; case 1: values[0] = __Pyx_ArgRef_FASTCALL(__pyx_args, 0); if (!CYTHON_ASSUME_SAFE_MACROS && unlikely(!values[0])) __PYX_ERR(0, 220, __pyx_L3_error) break; default: goto __pyx_L5_argtuple_error; } if (!values[1]) values[1] = __Pyx_NewRef(((PyObject *)((PyObject*)__pyx_mstate_global->__pyx_int_25))); } __pyx_v_data = values[0]; __pyx_v_sample_size = values[1]; } goto __pyx_L6_skip; __pyx_L5_argtuple_error:; __Pyx_RaiseArgtupleInvalid("get_jsonl_schema", 0, 1, 2, __pyx_nargs); __PYX_ERR(0, 220, __pyx_L3_error) __pyx_L6_skip:; goto __pyx_L4_argument_unpacking_done; __pyx_L3_error:; for (Py_ssize_t __pyx_temp=0; __pyx_temp < (Py_ssize_t)(sizeof(values)/sizeof(values[0])); ++__pyx_temp) { Py_XDECREF(values[__pyx_temp]); } __Pyx_AddTraceback("rugo.jsonl.get_jsonl_schema", __pyx_clineno, __pyx_lineno, __pyx_filename); __Pyx_RefNannyFinishContext(); return NULL; __pyx_L4_argument_unpacking_done:; __pyx_r = __pyx_pf_4rugo_5jsonl_2get_jsonl_schema(__pyx_self, __pyx_v_data, __pyx_v_sample_size); int __pyx_lineno = 0; const char *__pyx_filename = NULL; int __pyx_clineno = 0; /* function exit code */ for (Py_ssize_t __pyx_temp=0; __pyx_temp < (Py_ssize_t)(sizeof(values)/sizeof(values[0])); ++__pyx_temp) { Py_XDECREF(values[__pyx_temp]); } __Pyx_RefNannyFinishContext(); return __pyx_r; } static PyObject *__pyx_pf_4rugo_5jsonl_2get_jsonl_schema(CYTHON_UNUSED PyObject *__pyx_self, PyObject *__pyx_v_data, PyObject *__pyx_v_sample_size) { uint8_t const *__pyx_v_data_ptr; size_t __pyx_v_data_size; PyObject *__pyx_v_data_bytes = 0; Py_buffer __pyx_v_view; int __pyx_v_have_view; std::vector<ColumnSchema> __pyx_v_schema; PyObject *__pyx_v_result = NULL; size_t __pyx_v_i; int __pyx_v_type_val; ColumnSchema __pyx_v_col; PyObject *__pyx_v_type_str = NULL; int __pyx_v_elem_val; PyObject *__pyx_r = NULL; /* … */ /* function exit code */ __pyx_L1_error:; __Pyx_XDECREF(__pyx_t_2); __Pyx_XDECREF(__pyx_t_6); __Pyx_XDECREF(__pyx_t_7); __Pyx_AddTraceback("rugo.jsonl.get_jsonl_schema", __pyx_clineno, __pyx_lineno, __pyx_filename); __pyx_r = NULL; __pyx_L0:; __Pyx_XDECREF(__pyx_v_data_bytes); __Pyx_XDECREF(__pyx_v_result); __Pyx_XDECREF(__pyx_v_type_str); __Pyx_XGIVEREF(__pyx_r); __Pyx_RefNannyFinishContext(); return __pyx_r; } /* … */ __pyx_t_2 = __Pyx_CyFunction_New(&__pyx_mdef_4rugo_5jsonl_3get_jsonl_schema, 0, __pyx_mstate_global->__pyx_n_u_get_jsonl_schema, NULL, __pyx_mstate_global->__pyx_n_u_rugo_jsonl, __pyx_mstate_global->__pyx_d, ((PyObject *)__pyx_mstate_global->__pyx_codeobj_tab[5])); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 220, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_2); __Pyx_CyFunction_SetDefaultsTuple(__pyx_t_2, __pyx_mstate_global->__pyx_tuple[0]); if (PyDict_SetItem(__pyx_mstate_global->__pyx_d, __pyx_mstate_global->__pyx_n_u_get_jsonl_schema, __pyx_t_2) < (0)) __PYX_ERR(0, 220, __pyx_L1_error) __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; /* … */ __pyx_mstate_global->__pyx_tuple[0] = PyTuple_Pack(1, ((PyObject*)__pyx_mstate_global->__pyx_int_25)); if (unlikely(!__pyx_mstate_global->__pyx_tuple[0])) __PYX_ERR(0, 220, __pyx_L1_error) __Pyx_GOTREF(__pyx_mstate_global->__pyx_tuple[0]); __Pyx_GIVEREF(__pyx_mstate_global->__pyx_tuple[0]);
221: """
222: Infer the schema of a JSONL dataset from a sample of the data.
223:
224: Parameters
225: ----------
226: data : bytes or object supporting the buffer protocol
227: The JSONL data to analyze.
228: sample_size : int, optional
229: The number of rows to sample for schema inference (default: 25).
230:
231: Returns
232: -------
233: list of dict
234: A list of dictionaries, each describing a column with keys:
235: - 'name': str, the column name
236: - 'type': str, the inferred type ('null', 'boolean', 'int64', 'double', 'bytes')
237: - 'nullable': bool, whether the column can contain null values
238: """
239: cdef const uint8_t* data_ptr
240: cdef size_t data_size
241: cdef bytes data_bytes
242: cdef Py_buffer view
+243: cdef bint have_view = False
__pyx_v_have_view = 0;
+244: if isinstance(data, bytes):
__pyx_t_1 = PyBytes_Check(__pyx_v_data);
if (__pyx_t_1) {
/* … */
goto __pyx_L3;
}
+245: data_bytes = <bytes>data
__pyx_t_2 = __pyx_v_data;
__Pyx_INCREF(__pyx_t_2);
__pyx_v_data_bytes = ((PyObject*)__pyx_t_2);
__pyx_t_2 = 0;
+246: data_ptr = <const uint8_t*>(<char*>data_bytes)
if (unlikely(__pyx_v_data_bytes == Py_None)) {
PyErr_SetString(PyExc_TypeError, "expected bytes, NoneType found");
__PYX_ERR(0, 246, __pyx_L1_error)
}
__pyx_t_3 = __Pyx_PyBytes_AsWritableString(__pyx_v_data_bytes); if (unlikely((!__pyx_t_3) && PyErr_Occurred())) __PYX_ERR(0, 246, __pyx_L1_error)
__pyx_v_data_ptr = ((uint8_t const *)((char *)__pyx_t_3));
+247: data_size = len(data_bytes)
if (unlikely(__pyx_v_data_bytes == Py_None)) {
PyErr_SetString(PyExc_TypeError, "object of type 'NoneType' has no len()");
__PYX_ERR(0, 247, __pyx_L1_error)
}
__pyx_t_4 = __Pyx_PyBytes_GET_SIZE(__pyx_v_data_bytes); if (unlikely(__pyx_t_4 == ((Py_ssize_t)-1))) __PYX_ERR(0, 247, __pyx_L1_error)
__pyx_v_data_size = __pyx_t_4;
248: else:
+249: if PyObject_GetBuffer(data, &view, PyBUF_CONTIG_RO) == -1:
/*else*/ {
__pyx_t_5 = PyObject_GetBuffer(__pyx_v_data, (&__pyx_v_view), PyBUF_CONTIG_RO); if (unlikely(__pyx_t_5 == ((int)-1))) __PYX_ERR(0, 249, __pyx_L1_error)
__pyx_t_1 = (__pyx_t_5 == -1L);
if (unlikely(__pyx_t_1)) {
/* … */
}
+250: raise TypeError("object does not support contiguous buffer interface")
__pyx_t_6 = NULL;
__Pyx_INCREF(__pyx_builtin_TypeError);
__pyx_t_7 = __pyx_builtin_TypeError;
__pyx_t_8 = 1;
{
PyObject *__pyx_callargs[2] = {__pyx_t_6, __pyx_mstate_global->__pyx_kp_u_object_does_not_support_contiguo};
__pyx_t_2 = __Pyx_PyObject_FastCall(__pyx_t_7, __pyx_callargs+__pyx_t_8, (2-__pyx_t_8) | (__pyx_t_8*__Pyx_PY_VECTORCALL_ARGUMENTS_OFFSET));
__Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
__Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 250, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_2);
}
__Pyx_Raise(__pyx_t_2, 0, 0, 0);
__Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
__PYX_ERR(0, 250, __pyx_L1_error)
+251: have_view = True
__pyx_v_have_view = 1;
+252: data_ptr = <const uint8_t*>view.buf
__pyx_v_data_ptr = ((uint8_t const *)__pyx_v_view.buf);
+253: data_size = <size_t>view.len
__pyx_v_data_size = ((size_t)__pyx_v_view.len); } __pyx_L3:;
+254: cdef vector[ColumnSchema] schema = GetJsonlSchema(data_ptr, data_size, sample_size)
__pyx_t_8 = __Pyx_PyLong_As_size_t(__pyx_v_sample_size); if (unlikely((__pyx_t_8 == (size_t)-1) && PyErr_Occurred())) __PYX_ERR(0, 254, __pyx_L1_error) try { __pyx_t_9 = GetJsonlSchema(__pyx_v_data_ptr, __pyx_v_data_size, __pyx_t_8); } catch(...) { __Pyx_CppExn2PyErr(); __PYX_ERR(0, 254, __pyx_L1_error) } __pyx_v_schema = __PYX_STD_MOVE_IF_SUPPORTED(__pyx_t_9);
+255: if have_view:
if (__pyx_v_have_view) {
/* … */
}
+256: PyBuffer_Release(&view)
PyBuffer_Release((&__pyx_v_view));
+257: result = []
__pyx_t_2 = PyList_New(0); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 257, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_2); __pyx_v_result = ((PyObject*)__pyx_t_2); __pyx_t_2 = 0;
258: cdef size_t i
259: cdef int type_val
+260: for i in range(schema.size()):
__pyx_t_10 = __pyx_v_schema.size();
__pyx_t_11 = __pyx_t_10;
for (__pyx_t_8 = 0; __pyx_t_8 < __pyx_t_11; __pyx_t_8+=1) {
__pyx_v_i = __pyx_t_8;
+261: col = schema[i]
__pyx_v_col = (__pyx_v_schema[__pyx_v_i]);
+262: type_val = <int>col.type
__pyx_v_type_val = ((int)__pyx_v_col.type);
+263: type_str = "string"
__Pyx_INCREF(__pyx_mstate_global->__pyx_n_u_string); __Pyx_XDECREF_SET(__pyx_v_type_str, __pyx_mstate_global->__pyx_n_u_string);
264: # JsonType enum: Null=0, Boolean=1, Integer=2, Double=3, String=4, Array=5, Object=6
+265: if type_val == 0:
switch (__pyx_v_type_val) {
case 0:
/* … */
break;
case 1:
+266: type_str = "null"
__Pyx_INCREF(__pyx_mstate_global->__pyx_n_u_null); __Pyx_DECREF_SET(__pyx_v_type_str, __pyx_mstate_global->__pyx_n_u_null);
+267: elif type_val == 1:
break;
case 2:
+268: type_str = "boolean"
__Pyx_INCREF(__pyx_mstate_global->__pyx_n_u_boolean); __Pyx_DECREF_SET(__pyx_v_type_str, __pyx_mstate_global->__pyx_n_u_boolean);
+269: elif type_val == 2:
break;
case 3:
+270: type_str = "int64"
__Pyx_INCREF(__pyx_mstate_global->__pyx_n_u_int64); __Pyx_DECREF_SET(__pyx_v_type_str, __pyx_mstate_global->__pyx_n_u_int64);
+271: elif type_val == 3:
break;
case 4:
+272: type_str = "double"
__Pyx_INCREF(__pyx_mstate_global->__pyx_n_u_double); __Pyx_DECREF_SET(__pyx_v_type_str, __pyx_mstate_global->__pyx_n_u_double);
+273: elif type_val == 4:
break;
case 5:
+274: type_str = "string"
__Pyx_INCREF(__pyx_mstate_global->__pyx_n_u_string); __Pyx_DECREF_SET(__pyx_v_type_str, __pyx_mstate_global->__pyx_n_u_string);
+275: elif type_val == 5:
break;
case 6:
276: # array: include element type if available
+277: elem_val = <int>col.element_type
__pyx_v_elem_val = ((int)__pyx_v_col.element_type);
+278: if elem_val == 2:
switch (__pyx_v_elem_val) {
case 2:
/* … */
break;
case 3:
+279: type_str = "array<int64>"
__Pyx_INCREF(__pyx_mstate_global->__pyx_kp_u_array_int64); __Pyx_DECREF_SET(__pyx_v_type_str, __pyx_mstate_global->__pyx_kp_u_array_int64);
+280: elif elem_val == 3:
break;
case 4:
+281: type_str = "array<double>"
__Pyx_INCREF(__pyx_mstate_global->__pyx_kp_u_array_double); __Pyx_DECREF_SET(__pyx_v_type_str, __pyx_mstate_global->__pyx_kp_u_array_double);
+282: elif elem_val == 4:
break;
default:
+283: type_str = "array<bytes>"
__Pyx_INCREF(__pyx_mstate_global->__pyx_kp_u_array_bytes); __Pyx_DECREF_SET(__pyx_v_type_str, __pyx_mstate_global->__pyx_kp_u_array_bytes);
284: else:
+285: type_str = "array"
__Pyx_INCREF(__pyx_mstate_global->__pyx_n_u_array); __Pyx_DECREF_SET(__pyx_v_type_str, __pyx_mstate_global->__pyx_n_u_array); break; }
+286: elif type_val == 6:
break;
default: break;
}
+287: type_str = "object"
__Pyx_INCREF(__pyx_mstate_global->__pyx_n_u_object); __Pyx_DECREF_SET(__pyx_v_type_str, __pyx_mstate_global->__pyx_n_u_object);
+288: result.append({
__pyx_t_12 = __Pyx_PyList_Append(__pyx_v_result, __pyx_t_2); if (unlikely(__pyx_t_12 == ((int)-1))) __PYX_ERR(0, 288, __pyx_L1_error) __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; }
+289: 'name': col.name.decode('utf-8'),
__pyx_t_2 = __Pyx_PyDict_NewPresized(3); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 289, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_2); __pyx_t_7 = __Pyx_decode_cpp_string(__pyx_v_col.name, 0, PY_SSIZE_T_MAX, NULL, NULL, PyUnicode_DecodeUTF8); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 289, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_7); if (PyDict_SetItem(__pyx_t_2, __pyx_mstate_global->__pyx_n_u_name, __pyx_t_7) < (0)) __PYX_ERR(0, 289, __pyx_L1_error) __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
+290: 'type': type_str,
if (PyDict_SetItem(__pyx_t_2, __pyx_mstate_global->__pyx_n_u_type, __pyx_v_type_str) < (0)) __PYX_ERR(0, 289, __pyx_L1_error)
+291: 'nullable': col.nullable
__pyx_t_7 = __Pyx_PyBool_FromLong(__pyx_v_col.nullable); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 291, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_7); if (PyDict_SetItem(__pyx_t_2, __pyx_mstate_global->__pyx_n_u_nullable, __pyx_t_7) < (0)) __PYX_ERR(0, 289, __pyx_L1_error) __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
292: })
+293: return result
__Pyx_XDECREF(__pyx_r); __Pyx_INCREF(__pyx_v_result); __pyx_r = __pyx_v_result; goto __pyx_L0;
294:
295:
+296: def read_jsonl(data, columns=None, parse_arrays=True, parse_objects=True):
/* Python wrapper */ static PyObject *__pyx_pw_4rugo_5jsonl_5read_jsonl(PyObject *__pyx_self, #if CYTHON_METH_FASTCALL PyObject *const *__pyx_args, Py_ssize_t __pyx_nargs, PyObject *__pyx_kwds #else PyObject *__pyx_args, PyObject *__pyx_kwds #endif ); /*proto*/ PyDoc_STRVAR(__pyx_doc_4rugo_5jsonl_4read_jsonl, "\n Reads a JSONL (JSON Lines) dataset and returns its contents in a columnar format.\n\n Parameters\n ----------\n data : bytes or object supporting buffer protocol\n The JSONL data to read. Can be a bytes object or any object supporting the buffer protocol.\n columns : list of str, optional\n List of column names to read. If None, all columns are read.\n\n Returns\n -------\n dict\n A dictionary with the following keys:\n - 'success': bool, True if reading was successful.\n - 'column_names': list of str, names of the columns.\n - 'num_rows': int, number of rows in the dataset.\n - 'columns': list, each element is a list of values for a column (with None for nulls), or None if the column failed to read.\n "); static PyMethodDef __pyx_mdef_4rugo_5jsonl_5read_jsonl = {"read_jsonl", (PyCFunction)(void(*)(void))(__Pyx_PyCFunction_FastCallWithKeywords)__pyx_pw_4rugo_5jsonl_5read_jsonl, __Pyx_METH_FASTCALL|METH_KEYWORDS, __pyx_doc_4rugo_5jsonl_4read_jsonl}; static PyObject *__pyx_pw_4rugo_5jsonl_5read_jsonl(PyObject *__pyx_self, #if CYTHON_METH_FASTCALL PyObject *const *__pyx_args, Py_ssize_t __pyx_nargs, PyObject *__pyx_kwds #else PyObject *__pyx_args, PyObject *__pyx_kwds #endif ) { PyObject *__pyx_v_data = 0; PyObject *__pyx_v_columns = 0; PyObject *__pyx_v_parse_arrays = 0; CYTHON_UNUSED PyObject *__pyx_v_parse_objects = 0; #if !CYTHON_METH_FASTCALL CYTHON_UNUSED Py_ssize_t __pyx_nargs; #endif CYTHON_UNUSED PyObject *const *__pyx_kwvalues; PyObject *__pyx_r = 0; __Pyx_RefNannyDeclarations __Pyx_RefNannySetupContext("read_jsonl (wrapper)", 0); #if !CYTHON_METH_FASTCALL #if CYTHON_ASSUME_SAFE_SIZE __pyx_nargs = PyTuple_GET_SIZE(__pyx_args); #else __pyx_nargs = PyTuple_Size(__pyx_args); if (unlikely(__pyx_nargs < 0)) return NULL; #endif #endif __pyx_kwvalues = __Pyx_KwValues_FASTCALL(__pyx_args, __pyx_nargs); { PyObject ** const __pyx_pyargnames[] = {&__pyx_mstate_global->__pyx_n_u_data,&__pyx_mstate_global->__pyx_n_u_columns,&__pyx_mstate_global->__pyx_n_u_parse_arrays,&__pyx_mstate_global->__pyx_n_u_parse_objects,0}; PyObject* values[4] = {0,0,0,0}; const Py_ssize_t __pyx_kwds_len = (__pyx_kwds) ? __Pyx_NumKwargs_FASTCALL(__pyx_kwds) : 0; if (unlikely(__pyx_kwds_len) < 0) __PYX_ERR(0, 296, __pyx_L3_error) if (__pyx_kwds_len > 0) { switch (__pyx_nargs) { case 4: values[3] = __Pyx_ArgRef_FASTCALL(__pyx_args, 3); if (!CYTHON_ASSUME_SAFE_MACROS && unlikely(!values[3])) __PYX_ERR(0, 296, __pyx_L3_error) CYTHON_FALLTHROUGH; case 3: values[2] = __Pyx_ArgRef_FASTCALL(__pyx_args, 2); if (!CYTHON_ASSUME_SAFE_MACROS && unlikely(!values[2])) __PYX_ERR(0, 296, __pyx_L3_error) CYTHON_FALLTHROUGH; case 2: values[1] = __Pyx_ArgRef_FASTCALL(__pyx_args, 1); if (!CYTHON_ASSUME_SAFE_MACROS && unlikely(!values[1])) __PYX_ERR(0, 296, __pyx_L3_error) CYTHON_FALLTHROUGH; case 1: values[0] = __Pyx_ArgRef_FASTCALL(__pyx_args, 0); if (!CYTHON_ASSUME_SAFE_MACROS && unlikely(!values[0])) __PYX_ERR(0, 296, __pyx_L3_error) CYTHON_FALLTHROUGH; case 0: break; default: goto __pyx_L5_argtuple_error; } const Py_ssize_t kwd_pos_args = __pyx_nargs; if (__Pyx_ParseKeywords(__pyx_kwds, __pyx_kwvalues, __pyx_pyargnames, 0, values, kwd_pos_args, __pyx_kwds_len, "read_jsonl", 0) < (0)) __PYX_ERR(0, 296, __pyx_L3_error) if (!values[1]) values[1] = __Pyx_NewRef(((PyObject *)Py_None)); if (!values[2]) values[2] = __Pyx_NewRef(((PyObject *)((PyObject*)Py_True))); if (!values[3]) values[3] = __Pyx_NewRef(((PyObject *)((PyObject*)Py_True))); for (Py_ssize_t i = __pyx_nargs; i < 1; i++) { if (unlikely(!values[i])) { __Pyx_RaiseArgtupleInvalid("read_jsonl", 0, 1, 4, i); __PYX_ERR(0, 296, __pyx_L3_error) } } } else { switch (__pyx_nargs) { case 4: values[3] = __Pyx_ArgRef_FASTCALL(__pyx_args, 3); if (!CYTHON_ASSUME_SAFE_MACROS && unlikely(!values[3])) __PYX_ERR(0, 296, __pyx_L3_error) CYTHON_FALLTHROUGH; case 3: values[2] = __Pyx_ArgRef_FASTCALL(__pyx_args, 2); if (!CYTHON_ASSUME_SAFE_MACROS && unlikely(!values[2])) __PYX_ERR(0, 296, __pyx_L3_error) CYTHON_FALLTHROUGH; case 2: values[1] = __Pyx_ArgRef_FASTCALL(__pyx_args, 1); if (!CYTHON_ASSUME_SAFE_MACROS && unlikely(!values[1])) __PYX_ERR(0, 296, __pyx_L3_error) CYTHON_FALLTHROUGH; case 1: values[0] = __Pyx_ArgRef_FASTCALL(__pyx_args, 0); if (!CYTHON_ASSUME_SAFE_MACROS && unlikely(!values[0])) __PYX_ERR(0, 296, __pyx_L3_error) break; default: goto __pyx_L5_argtuple_error; } if (!values[1]) values[1] = __Pyx_NewRef(((PyObject *)Py_None)); if (!values[2]) values[2] = __Pyx_NewRef(((PyObject *)((PyObject*)Py_True))); if (!values[3]) values[3] = __Pyx_NewRef(((PyObject *)((PyObject*)Py_True))); } __pyx_v_data = values[0]; __pyx_v_columns = values[1]; __pyx_v_parse_arrays = values[2]; __pyx_v_parse_objects = values[3]; } goto __pyx_L6_skip; __pyx_L5_argtuple_error:; __Pyx_RaiseArgtupleInvalid("read_jsonl", 0, 1, 4, __pyx_nargs); __PYX_ERR(0, 296, __pyx_L3_error) __pyx_L6_skip:; goto __pyx_L4_argument_unpacking_done; __pyx_L3_error:; for (Py_ssize_t __pyx_temp=0; __pyx_temp < (Py_ssize_t)(sizeof(values)/sizeof(values[0])); ++__pyx_temp) { Py_XDECREF(values[__pyx_temp]); } __Pyx_AddTraceback("rugo.jsonl.read_jsonl", __pyx_clineno, __pyx_lineno, __pyx_filename); __Pyx_RefNannyFinishContext(); return NULL; __pyx_L4_argument_unpacking_done:; __pyx_r = __pyx_pf_4rugo_5jsonl_4read_jsonl(__pyx_self, __pyx_v_data, __pyx_v_columns, __pyx_v_parse_arrays, __pyx_v_parse_objects); /* function exit code */ for (Py_ssize_t __pyx_temp=0; __pyx_temp < (Py_ssize_t)(sizeof(values)/sizeof(values[0])); ++__pyx_temp) { Py_XDECREF(values[__pyx_temp]); } __Pyx_RefNannyFinishContext(); return __pyx_r; } /* … */ static PyObject *__pyx_pf_4rugo_5jsonl_4read_jsonl(CYTHON_UNUSED PyObject *__pyx_self, PyObject *__pyx_v_data, PyObject *__pyx_v_columns, PyObject *__pyx_v_parse_arrays, CYTHON_UNUSED PyObject *__pyx_v_parse_objects) { struct __pyx_obj_4rugo_5jsonl___pyx_scope_struct_1_read_jsonl *__pyx_cur_scope; uint8_t const *__pyx_v_data_ptr; size_t __pyx_v_data_size; PyObject *__pyx_v_data_bytes = 0; Py_buffer __pyx_v_view; int __pyx_v_have_view; std::vector<std::string> __pyx_v_column_names_vec; JsonlTable __pyx_v_table; PyObject *__pyx_v_col_name = NULL; PyObject *__pyx_v_py_column_names = NULL; size_t __pyx_v_i; PyObject *__pyx_v_py_columns = NULL; PyObject *__pyx_v_o = NULL; PyObject *__pyx_v_o_ptr; JsonlColumn *__pyx_v_col; PyObject *__pyx_v_col_type = NULL; PyObject *__pyx_v_py_list = NULL; std::vector<std::string> ::size_type __pyx_v_j; std::string __pyx_v_raw; PyObject *__pyx_v_py_obj = NULL; PyObject *__pyx_v_elem_type = NULL; PyObject *__pyx_v_parsed = NULL; char __pyx_v_first; PyObject *__pyx_r = NULL; __pyx_cur_scope = (struct __pyx_obj_4rugo_5jsonl___pyx_scope_struct_1_read_jsonl *)__pyx_tp_new_4rugo_5jsonl___pyx_scope_struct_1_read_jsonl(__pyx_mstate_global->__pyx_ptype_4rugo_5jsonl___pyx_scope_struct_1_read_jsonl, __pyx_mstate_global->__pyx_empty_tuple, NULL); if (unlikely(!__pyx_cur_scope)) { __pyx_cur_scope = ((struct __pyx_obj_4rugo_5jsonl___pyx_scope_struct_1_read_jsonl *)Py_None); __Pyx_INCREF(Py_None); __PYX_ERR(0, 296, __pyx_L1_error) } else { __Pyx_GOTREF((PyObject *)__pyx_cur_scope); } /* … */ __pyx_t_2 = __Pyx_CyFunction_New(&__pyx_mdef_4rugo_5jsonl_5read_jsonl, 0, __pyx_mstate_global->__pyx_n_u_read_jsonl, NULL, __pyx_mstate_global->__pyx_n_u_rugo_jsonl, __pyx_mstate_global->__pyx_d, ((PyObject *)__pyx_mstate_global->__pyx_codeobj_tab[6])); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 296, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_2); __Pyx_CyFunction_SetDefaultsTuple(__pyx_t_2, __pyx_mstate_global->__pyx_tuple[1]); if (PyDict_SetItem(__pyx_mstate_global->__pyx_d, __pyx_mstate_global->__pyx_n_u_read_jsonl, __pyx_t_2) < (0)) __PYX_ERR(0, 296, __pyx_L1_error) __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
297: """
298: Reads a JSONL (JSON Lines) dataset and returns its contents in a columnar format.
299:
300: Parameters
301: ----------
302: data : bytes or object supporting buffer protocol
303: The JSONL data to read. Can be a bytes object or any object supporting the buffer protocol.
304: columns : list of str, optional
305: List of column names to read. If None, all columns are read.
306:
307: Returns
308: -------
309: dict
310: A dictionary with the following keys:
311: - 'success': bool, True if reading was successful.
312: - 'column_names': list of str, names of the columns.
313: - 'num_rows': int, number of rows in the dataset.
314: - 'columns': list, each element is a list of values for a column (with None for nulls), or None if the column failed to read.
315: """
316: cdef const uint8_t* data_ptr
317: cdef size_t data_size
318: cdef bytes data_bytes
319: cdef Py_buffer view
+320: cdef bint have_view = False
__pyx_v_have_view = 0;
+321: if isinstance(data, bytes):
__pyx_t_1 = PyBytes_Check(__pyx_v_data);
if (__pyx_t_1) {
/* … */
goto __pyx_L3;
}
+322: data_bytes = <bytes>data
__pyx_t_2 = __pyx_v_data;
__Pyx_INCREF(__pyx_t_2);
__pyx_v_data_bytes = ((PyObject*)__pyx_t_2);
__pyx_t_2 = 0;
+323: data_ptr = <const uint8_t*>(<char*>data_bytes)
if (unlikely(__pyx_v_data_bytes == Py_None)) {
PyErr_SetString(PyExc_TypeError, "expected bytes, NoneType found");
__PYX_ERR(0, 323, __pyx_L1_error)
}
__pyx_t_3 = __Pyx_PyBytes_AsWritableString(__pyx_v_data_bytes); if (unlikely((!__pyx_t_3) && PyErr_Occurred())) __PYX_ERR(0, 323, __pyx_L1_error)
__pyx_v_data_ptr = ((uint8_t const *)((char *)__pyx_t_3));
+324: data_size = len(data_bytes)
if (unlikely(__pyx_v_data_bytes == Py_None)) {
PyErr_SetString(PyExc_TypeError, "object of type 'NoneType' has no len()");
__PYX_ERR(0, 324, __pyx_L1_error)
}
__pyx_t_4 = __Pyx_PyBytes_GET_SIZE(__pyx_v_data_bytes); if (unlikely(__pyx_t_4 == ((Py_ssize_t)-1))) __PYX_ERR(0, 324, __pyx_L1_error)
__pyx_v_data_size = __pyx_t_4;
325: else:
+326: if PyObject_GetBuffer(data, &view, PyBUF_CONTIG_RO) == -1:
/*else*/ {
__pyx_t_5 = PyObject_GetBuffer(__pyx_v_data, (&__pyx_v_view), PyBUF_CONTIG_RO); if (unlikely(__pyx_t_5 == ((int)-1))) __PYX_ERR(0, 326, __pyx_L1_error)
__pyx_t_1 = (__pyx_t_5 == -1L);
if (unlikely(__pyx_t_1)) {
/* … */
}
+327: raise TypeError("object does not support contiguous buffer interface")
__pyx_t_6 = NULL;
__Pyx_INCREF(__pyx_builtin_TypeError);
__pyx_t_7 = __pyx_builtin_TypeError;
__pyx_t_8 = 1;
{
PyObject *__pyx_callargs[2] = {__pyx_t_6, __pyx_mstate_global->__pyx_kp_u_object_does_not_support_contiguo};
__pyx_t_2 = __Pyx_PyObject_FastCall(__pyx_t_7, __pyx_callargs+__pyx_t_8, (2-__pyx_t_8) | (__pyx_t_8*__Pyx_PY_VECTORCALL_ARGUMENTS_OFFSET));
__Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
__Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 327, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_2);
}
__Pyx_Raise(__pyx_t_2, 0, 0, 0);
__Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
__PYX_ERR(0, 327, __pyx_L1_error)
+328: have_view = True
__pyx_v_have_view = 1;
+329: data_ptr = <const uint8_t*>view.buf
__pyx_v_data_ptr = ((uint8_t const *)__pyx_v_view.buf);
+330: data_size = <size_t>view.len
__pyx_v_data_size = ((size_t)__pyx_v_view.len); } __pyx_L3:;
331: cdef vector[string] column_names_vec
332: cdef JsonlTable table
+333: if columns is None:
__pyx_t_1 = (__pyx_v_columns == Py_None);
if (__pyx_t_1) {
/* … */
goto __pyx_L5;
}
+334: table = ReadJsonl(data_ptr, data_size)
try {
__pyx_t_9 = ReadJsonl(__pyx_v_data_ptr, __pyx_v_data_size);
} catch(...) {
__Pyx_CppExn2PyErr();
__PYX_ERR(0, 334, __pyx_L1_error)
}
__pyx_v_table = __PYX_STD_MOVE_IF_SUPPORTED(__pyx_t_9);
335: else:
+336: for col_name in columns:
/*else*/ {
if (likely(PyList_CheckExact(__pyx_v_columns)) || PyTuple_CheckExact(__pyx_v_columns)) {
__pyx_t_2 = __pyx_v_columns; __Pyx_INCREF(__pyx_t_2);
__pyx_t_4 = 0;
__pyx_t_10 = NULL;
} else {
__pyx_t_4 = -1; __pyx_t_2 = PyObject_GetIter(__pyx_v_columns); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 336, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_2);
__pyx_t_10 = (CYTHON_COMPILING_IN_LIMITED_API) ? PyIter_Next : __Pyx_PyObject_GetIterNextFunc(__pyx_t_2); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 336, __pyx_L1_error)
}
for (;;) {
if (likely(!__pyx_t_10)) {
if (likely(PyList_CheckExact(__pyx_t_2))) {
{
Py_ssize_t __pyx_temp = __Pyx_PyList_GET_SIZE(__pyx_t_2);
#if !CYTHON_ASSUME_SAFE_SIZE
if (unlikely((__pyx_temp < 0))) __PYX_ERR(0, 336, __pyx_L1_error)
#endif
if (__pyx_t_4 >= __pyx_temp) break;
}
__pyx_t_7 = __Pyx_PyList_GetItemRef(__pyx_t_2, __pyx_t_4);
++__pyx_t_4;
} else {
{
Py_ssize_t __pyx_temp = __Pyx_PyTuple_GET_SIZE(__pyx_t_2);
#if !CYTHON_ASSUME_SAFE_SIZE
if (unlikely((__pyx_temp < 0))) __PYX_ERR(0, 336, __pyx_L1_error)
#endif
if (__pyx_t_4 >= __pyx_temp) break;
}
#if CYTHON_ASSUME_SAFE_MACROS && !CYTHON_AVOID_BORROWED_REFS
__pyx_t_7 = __Pyx_NewRef(PyTuple_GET_ITEM(__pyx_t_2, __pyx_t_4));
#else
__pyx_t_7 = __Pyx_PySequence_ITEM(__pyx_t_2, __pyx_t_4);
#endif
++__pyx_t_4;
}
if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 336, __pyx_L1_error)
} else {
__pyx_t_7 = __pyx_t_10(__pyx_t_2);
if (unlikely(!__pyx_t_7)) {
PyObject* exc_type = PyErr_Occurred();
if (exc_type) {
if (unlikely(!__Pyx_PyErr_GivenExceptionMatches(exc_type, PyExc_StopIteration))) __PYX_ERR(0, 336, __pyx_L1_error)
PyErr_Clear();
}
break;
}
}
__Pyx_GOTREF(__pyx_t_7);
__Pyx_XDECREF_SET(__pyx_v_col_name, __pyx_t_7);
__pyx_t_7 = 0;
/* … */
}
__Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
+337: column_names_vec.push_back(col_name.encode('utf-8'))
__pyx_t_6 = __pyx_v_col_name;
__Pyx_INCREF(__pyx_t_6);
__pyx_t_8 = 0;
{
PyObject *__pyx_callargs[2] = {__pyx_t_6, __pyx_mstate_global->__pyx_kp_u_utf_8};
__pyx_t_7 = __Pyx_PyObject_FastCallMethod(__pyx_mstate_global->__pyx_n_u_encode, __pyx_callargs+__pyx_t_8, (2-__pyx_t_8) | (1*__Pyx_PY_VECTORCALL_ARGUMENTS_OFFSET));
__Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 337, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_7);
}
__pyx_t_11 = __pyx_convert_string_from_py_6libcpp_6string_std__in_string(__pyx_t_7); if (unlikely(PyErr_Occurred())) __PYX_ERR(0, 337, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
try {
__pyx_v_column_names_vec.push_back(__pyx_t_11);
} catch(...) {
__Pyx_CppExn2PyErr();
__PYX_ERR(0, 337, __pyx_L1_error)
}
+338: table = ReadJsonl(data_ptr, data_size, column_names_vec)
try {
__pyx_t_9 = ReadJsonl(__pyx_v_data_ptr, __pyx_v_data_size, __pyx_v_column_names_vec);
} catch(...) {
__Pyx_CppExn2PyErr();
__PYX_ERR(0, 338, __pyx_L1_error)
}
__pyx_v_table = __PYX_STD_MOVE_IF_SUPPORTED(__pyx_t_9);
}
__pyx_L5:;
+339: if have_view:
if (__pyx_v_have_view) {
/* … */
}
+340: PyBuffer_Release(&view)
PyBuffer_Release((&__pyx_v_view));
+341: if not table.success:
__pyx_t_1 = (!__pyx_v_table.success);
if (__pyx_t_1) {
/* … */
}
+342: return {
__Pyx_XDECREF(__pyx_r);
+343: 'success': False,
__pyx_t_2 = __Pyx_PyDict_NewPresized(4); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 343, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_2); if (PyDict_SetItem(__pyx_t_2, __pyx_mstate_global->__pyx_n_u_success, Py_False) < (0)) __PYX_ERR(0, 343, __pyx_L1_error)
+344: 'column_names': [],
__pyx_t_7 = PyList_New(0); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 344, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_7); if (PyDict_SetItem(__pyx_t_2, __pyx_mstate_global->__pyx_n_u_column_names, __pyx_t_7) < (0)) __PYX_ERR(0, 343, __pyx_L1_error) __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0; if (PyDict_SetItem(__pyx_t_2, __pyx_mstate_global->__pyx_n_u_num_rows, __pyx_mstate_global->__pyx_int_0) < (0)) __PYX_ERR(0, 343, __pyx_L1_error)
345: 'num_rows': 0,
+346: 'columns': []
__pyx_t_7 = PyList_New(0); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 346, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_7); if (PyDict_SetItem(__pyx_t_2, __pyx_mstate_global->__pyx_n_u_columns, __pyx_t_7) < (0)) __PYX_ERR(0, 343, __pyx_L1_error) __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0; __pyx_r = __pyx_t_2; __pyx_t_2 = 0; goto __pyx_L0;
347: }
+348: py_column_names = []
__pyx_t_2 = PyList_New(0); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 348, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_2); __pyx_v_py_column_names = ((PyObject*)__pyx_t_2); __pyx_t_2 = 0;
349: cdef size_t i
+350: for i in range(table.column_names.size()):
__pyx_t_12 = __pyx_v_table.column_names.size();
__pyx_t_13 = __pyx_t_12;
for (__pyx_t_8 = 0; __pyx_t_8 < __pyx_t_13; __pyx_t_8+=1) {
__pyx_v_i = __pyx_t_8;
+351: py_column_names.append(table.column_names[i].decode('utf-8'))
__pyx_t_2 = __Pyx_decode_cpp_string((__pyx_v_table.column_names[__pyx_v_i]), 0, PY_SSIZE_T_MAX, NULL, NULL, PyUnicode_DecodeUTF8); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 351, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_2); __pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_column_names, __pyx_t_2); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 351, __pyx_L1_error) __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; }
+352: py_columns = []
__pyx_t_2 = PyList_New(0); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 352, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_2); __pyx_v_py_columns = ((PyObject*)__pyx_t_2); __pyx_t_2 = 0;
353: # Python-level temporary for simdjson results
+354: o = None
__Pyx_INCREF(Py_None);
__pyx_v_o = Py_None;
355: cdef PyObject* o_ptr
356: cdef JsonlColumn* col
+357: for i in range(table.columns.size()):
__pyx_t_15 = __pyx_v_table.columns.size();
__pyx_t_16 = __pyx_t_15;
for (__pyx_t_8 = 0; __pyx_t_8 < __pyx_t_16; __pyx_t_8+=1) {
__pyx_v_i = __pyx_t_8;
+358: col = &table.columns[i]
__pyx_v_col = (&(__pyx_v_table.columns[__pyx_v_i]));
+359: if not col.success:
__pyx_t_1 = (!__pyx_v_col->success);
if (__pyx_t_1) {
/* … */
}
+360: py_columns.append(None)
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_columns, Py_None); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 360, __pyx_L1_error)
+361: continue
goto __pyx_L13_continue;
+362: col_type = col.type.decode('utf-8')
__pyx_t_2 = __Pyx_decode_cpp_string(__pyx_v_col->type, 0, PY_SSIZE_T_MAX, NULL, NULL, PyUnicode_DecodeUTF8); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 362, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_2); __Pyx_XDECREF_SET(__pyx_v_col_type, __pyx_t_2); __pyx_t_2 = 0;
+363: if col_type == 'int64':
__pyx_t_1 = (__Pyx_PyUnicode_Equals(__pyx_v_col_type, __pyx_mstate_global->__pyx_n_u_int64, Py_EQ)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 363, __pyx_L1_error) if (__pyx_t_1) { /* … */ goto __pyx_L16; }
+364: py_list = []
__pyx_t_2 = PyList_New(0); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 364, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_2); __Pyx_XDECREF_SET(__pyx_v_py_list, ((PyObject*)__pyx_t_2)); __pyx_t_2 = 0;
+365: for j in range(col.int_values.size()):
__pyx_t_17 = __pyx_v_col->int_values.size();
__pyx_t_18 = __pyx_t_17;
for (__pyx_t_12 = 0; __pyx_t_12 < __pyx_t_18; __pyx_t_12+=1) {
__pyx_v_j = __pyx_t_12;
+366: if col.null_mask[j]:
__pyx_t_1 = ((__pyx_v_col->null_mask[__pyx_v_j]) != 0);
if (__pyx_t_1) {
/* … */
goto __pyx_L19;
}
+367: py_list.append(None)
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_list, Py_None); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 367, __pyx_L1_error)
368: else:
+369: py_list.append(col.int_values[j])
/*else*/ {
__pyx_t_2 = __Pyx_PyLong_From_int64_t((__pyx_v_col->int_values[__pyx_v_j])); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 369, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_2);
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_list, __pyx_t_2); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 369, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
}
__pyx_L19:;
}
+370: py_columns.append(py_list)
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_columns, __pyx_v_py_list); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 370, __pyx_L1_error)
+371: elif col_type == 'double':
__pyx_t_1 = (__Pyx_PyUnicode_Equals(__pyx_v_col_type, __pyx_mstate_global->__pyx_n_u_double, Py_EQ)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 371, __pyx_L1_error) if (__pyx_t_1) { /* … */ goto __pyx_L16; }
+372: py_list = []
__pyx_t_2 = PyList_New(0); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 372, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_2); __Pyx_XDECREF_SET(__pyx_v_py_list, ((PyObject*)__pyx_t_2)); __pyx_t_2 = 0;
+373: for j in range(col.double_values.size()):
__pyx_t_19 = __pyx_v_col->double_values.size();
__pyx_t_20 = __pyx_t_19;
for (__pyx_t_12 = 0; __pyx_t_12 < __pyx_t_20; __pyx_t_12+=1) {
__pyx_v_j = __pyx_t_12;
+374: if col.null_mask[j]:
__pyx_t_1 = ((__pyx_v_col->null_mask[__pyx_v_j]) != 0);
if (__pyx_t_1) {
/* … */
goto __pyx_L22;
}
+375: py_list.append(None)
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_list, Py_None); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 375, __pyx_L1_error)
376: else:
+377: py_list.append(col.double_values[j])
/*else*/ {
__pyx_t_2 = PyFloat_FromDouble((__pyx_v_col->double_values[__pyx_v_j])); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 377, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_2);
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_list, __pyx_t_2); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 377, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
}
__pyx_L22:;
}
+378: py_columns.append(py_list)
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_columns, __pyx_v_py_list); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 378, __pyx_L1_error)
+379: elif col_type == 'string' or col_type == 'bytes':
__pyx_t_21 = (__Pyx_PyUnicode_Equals(__pyx_v_col_type, __pyx_mstate_global->__pyx_n_u_string, Py_EQ)); if (unlikely((__pyx_t_21 < 0))) __PYX_ERR(0, 379, __pyx_L1_error) if (!__pyx_t_21) { } else { __pyx_t_1 = __pyx_t_21; goto __pyx_L23_bool_binop_done; } __pyx_t_21 = (__Pyx_PyUnicode_Equals(__pyx_v_col_type, __pyx_mstate_global->__pyx_n_u_bytes, Py_EQ)); if (unlikely((__pyx_t_21 < 0))) __PYX_ERR(0, 379, __pyx_L1_error) __pyx_t_1 = __pyx_t_21; __pyx_L23_bool_binop_done:; if (__pyx_t_1) { /* … */ goto __pyx_L16; }
380: # Bytes columns: return as bytes (binary data), do NOT parse as JSON
381: # The schema has already determined this is a bytes/string column, not array/object
+382: py_list = []
__pyx_t_2 = PyList_New(0); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 382, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_2); __Pyx_XDECREF_SET(__pyx_v_py_list, ((PyObject*)__pyx_t_2)); __pyx_t_2 = 0;
+383: for j in range(col.string_values.size()):
__pyx_t_12 = __pyx_v_col->string_values.size();
__pyx_t_13 = __pyx_t_12;
for (__pyx_t_22 = 0; __pyx_t_22 < __pyx_t_13; __pyx_t_22+=1) {
__pyx_v_j = __pyx_t_22;
+384: if col.null_mask[j]:
__pyx_t_1 = ((__pyx_v_col->null_mask[__pyx_v_j]) != 0);
if (__pyx_t_1) {
/* … */
}
+385: py_list.append(None)
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_list, Py_None); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 385, __pyx_L1_error)
+386: continue
goto __pyx_L25_continue;
387:
+388: raw = col.string_values[j]
__pyx_v_raw = (__pyx_v_col->string_values[__pyx_v_j]);
+389: if raw.size() == 0:
__pyx_t_1 = (__pyx_v_raw.size() == 0);
if (__pyx_t_1) {
/* … */
}
+390: py_list.append(b'')
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_list, __pyx_mstate_global->__pyx_kp_b_); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 390, __pyx_L1_error)
+391: continue
goto __pyx_L25_continue;
392:
393: # Always return as bytes, never parse as JSON
+394: py_obj = <object>PyBytes_FromStringAndSize(<const char*>raw.data(), <Py_ssize_t>raw.size())
__pyx_t_2 = PyBytes_FromStringAndSize(((char const *)__pyx_v_raw.data()), ((Py_ssize_t)__pyx_v_raw.size())); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 394, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_2); __pyx_t_7 = __pyx_t_2; __Pyx_INCREF(__pyx_t_7); __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; __Pyx_XDECREF_SET(__pyx_v_py_obj, __pyx_t_7); __pyx_t_7 = 0;
+395: if py_obj is not None:
__pyx_t_1 = (__pyx_v_py_obj != Py_None);
if (__pyx_t_1) {
/* … */
goto __pyx_L29;
}
+396: py_list.append(py_obj)
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_list, __pyx_v_py_obj); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 396, __pyx_L1_error)
397: else:
+398: py_list.append(b'')
/*else*/ {
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_list, __pyx_mstate_global->__pyx_kp_b_); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 398, __pyx_L1_error)
}
__pyx_L29:;
__pyx_L25_continue:;
}
+399: py_columns.append(py_list)
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_columns, __pyx_v_py_list); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 399, __pyx_L1_error)
+400: elif col_type.startswith('array'):
__pyx_t_2 = __pyx_v_col_type;
__Pyx_INCREF(__pyx_t_2);
__pyx_t_23 = 0;
{
PyObject *__pyx_callargs[2] = {__pyx_t_2, __pyx_mstate_global->__pyx_n_u_array};
__pyx_t_7 = __Pyx_PyObject_FastCallMethod(__pyx_mstate_global->__pyx_n_u_startswith, __pyx_callargs+__pyx_t_23, (2-__pyx_t_23) | (1*__Pyx_PY_VECTORCALL_ARGUMENTS_OFFSET));
__Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 400, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_7);
}
__pyx_t_1 = __Pyx_PyObject_IsTrue(__pyx_t_7); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 400, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
if (__pyx_t_1) {
/* … */
goto __pyx_L16;
}
401: # Array columns: may be annotated as array<elemtype>
402: # Determine element type if provided (e.g. array<bytes>)
+403: elem_type = None
__Pyx_INCREF(Py_None); __Pyx_XDECREF_SET(__pyx_v_elem_type, Py_None);
+404: if col_type.startswith('array<') and col_type.endswith('>'):
__pyx_t_2 = __pyx_v_col_type;
__Pyx_INCREF(__pyx_t_2);
__pyx_t_23 = 0;
{
PyObject *__pyx_callargs[2] = {__pyx_t_2, __pyx_mstate_global->__pyx_kp_u_array_2};
__pyx_t_7 = __Pyx_PyObject_FastCallMethod(__pyx_mstate_global->__pyx_n_u_startswith, __pyx_callargs+__pyx_t_23, (2-__pyx_t_23) | (1*__Pyx_PY_VECTORCALL_ARGUMENTS_OFFSET));
__Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 404, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_7);
}
__pyx_t_21 = __Pyx_PyObject_IsTrue(__pyx_t_7); if (unlikely((__pyx_t_21 < 0))) __PYX_ERR(0, 404, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
if (__pyx_t_21) {
} else {
__pyx_t_1 = __pyx_t_21;
goto __pyx_L31_bool_binop_done;
}
__pyx_t_2 = __pyx_v_col_type;
__Pyx_INCREF(__pyx_t_2);
__pyx_t_23 = 0;
{
PyObject *__pyx_callargs[2] = {__pyx_t_2, __pyx_mstate_global->__pyx_kp_u__11};
__pyx_t_7 = __Pyx_PyObject_FastCallMethod(__pyx_mstate_global->__pyx_n_u_endswith, __pyx_callargs+__pyx_t_23, (2-__pyx_t_23) | (1*__Pyx_PY_VECTORCALL_ARGUMENTS_OFFSET));
__Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 404, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_7);
}
__pyx_t_21 = __Pyx_PyObject_IsTrue(__pyx_t_7); if (unlikely((__pyx_t_21 < 0))) __PYX_ERR(0, 404, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
__pyx_t_1 = __pyx_t_21;
__pyx_L31_bool_binop_done:;
if (__pyx_t_1) {
/* … */
}
+405: elem_type = col_type[6:-1]
__pyx_t_7 = __Pyx_PyObject_GetSlice(__pyx_v_col_type, 6, -1L, NULL, NULL, &__pyx_mstate_global->__pyx_slice[0], 1, 1, 0); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 405, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_7); __Pyx_DECREF_SET(__pyx_v_elem_type, __pyx_t_7); __pyx_t_7 = 0; /* … */ __pyx_mstate_global->__pyx_slice[0] = PySlice_New(__pyx_mstate_global->__pyx_int_6, __pyx_mstate_global->__pyx_int_neg_1, Py_None); if (unlikely(!__pyx_mstate_global->__pyx_slice[0])) __PYX_ERR(0, 405, __pyx_L1_error) __Pyx_GOTREF(__pyx_mstate_global->__pyx_slice[0]); __Pyx_GIVEREF(__pyx_mstate_global->__pyx_slice[0]);
406:
+407: def _convert_strings_to_bytes_inplace(obj):
/* Python wrapper */
static PyObject *__pyx_pw_4rugo_5jsonl_10read_jsonl_1_convert_strings_to_bytes_inplace(PyObject *__pyx_self,
#if CYTHON_METH_FASTCALL
PyObject *const *__pyx_args, Py_ssize_t __pyx_nargs, PyObject *__pyx_kwds
#else
PyObject *__pyx_args, PyObject *__pyx_kwds
#endif
); /*proto*/
static PyMethodDef __pyx_mdef_4rugo_5jsonl_10read_jsonl_1_convert_strings_to_bytes_inplace = {"_convert_strings_to_bytes_inplace", (PyCFunction)(void(*)(void))(__Pyx_PyCFunction_FastCallWithKeywords)__pyx_pw_4rugo_5jsonl_10read_jsonl_1_convert_strings_to_bytes_inplace, __Pyx_METH_FASTCALL|METH_KEYWORDS, 0};
static PyObject *__pyx_pw_4rugo_5jsonl_10read_jsonl_1_convert_strings_to_bytes_inplace(PyObject *__pyx_self,
#if CYTHON_METH_FASTCALL
PyObject *const *__pyx_args, Py_ssize_t __pyx_nargs, PyObject *__pyx_kwds
#else
PyObject *__pyx_args, PyObject *__pyx_kwds
#endif
) {
PyObject *__pyx_v_obj = 0;
#if !CYTHON_METH_FASTCALL
CYTHON_UNUSED Py_ssize_t __pyx_nargs;
#endif
CYTHON_UNUSED PyObject *const *__pyx_kwvalues;
PyObject *__pyx_r = 0;
__Pyx_RefNannyDeclarations
__Pyx_RefNannySetupContext("_convert_strings_to_bytes_inplace (wrapper)", 0);
#if !CYTHON_METH_FASTCALL
#if CYTHON_ASSUME_SAFE_SIZE
__pyx_nargs = PyTuple_GET_SIZE(__pyx_args);
#else
__pyx_nargs = PyTuple_Size(__pyx_args); if (unlikely(__pyx_nargs < 0)) return NULL;
#endif
#endif
__pyx_kwvalues = __Pyx_KwValues_FASTCALL(__pyx_args, __pyx_nargs);
{
PyObject ** const __pyx_pyargnames[] = {&__pyx_mstate_global->__pyx_n_u_obj,0};
PyObject* values[1] = {0};
const Py_ssize_t __pyx_kwds_len = (__pyx_kwds) ? __Pyx_NumKwargs_FASTCALL(__pyx_kwds) : 0;
if (unlikely(__pyx_kwds_len) < 0) __PYX_ERR(0, 407, __pyx_L3_error)
if (__pyx_kwds_len > 0) {
switch (__pyx_nargs) {
case 1:
values[0] = __Pyx_ArgRef_FASTCALL(__pyx_args, 0);
if (!CYTHON_ASSUME_SAFE_MACROS && unlikely(!values[0])) __PYX_ERR(0, 407, __pyx_L3_error)
CYTHON_FALLTHROUGH;
case 0: break;
default: goto __pyx_L5_argtuple_error;
}
const Py_ssize_t kwd_pos_args = __pyx_nargs;
if (__Pyx_ParseKeywords(__pyx_kwds, __pyx_kwvalues, __pyx_pyargnames, 0, values, kwd_pos_args, __pyx_kwds_len, "_convert_strings_to_bytes_inplace", 0) < (0)) __PYX_ERR(0, 407, __pyx_L3_error)
for (Py_ssize_t i = __pyx_nargs; i < 1; i++) {
if (unlikely(!values[i])) { __Pyx_RaiseArgtupleInvalid("_convert_strings_to_bytes_inplace", 1, 1, 1, i); __PYX_ERR(0, 407, __pyx_L3_error) }
}
} else if (unlikely(__pyx_nargs != 1)) {
goto __pyx_L5_argtuple_error;
} else {
values[0] = __Pyx_ArgRef_FASTCALL(__pyx_args, 0);
if (!CYTHON_ASSUME_SAFE_MACROS && unlikely(!values[0])) __PYX_ERR(0, 407, __pyx_L3_error)
}
__pyx_v_obj = values[0];
}
goto __pyx_L6_skip;
__pyx_L5_argtuple_error:;
__Pyx_RaiseArgtupleInvalid("_convert_strings_to_bytes_inplace", 1, 1, 1, __pyx_nargs); __PYX_ERR(0, 407, __pyx_L3_error)
__pyx_L6_skip:;
goto __pyx_L4_argument_unpacking_done;
__pyx_L3_error:;
for (Py_ssize_t __pyx_temp=0; __pyx_temp < (Py_ssize_t)(sizeof(values)/sizeof(values[0])); ++__pyx_temp) {
Py_XDECREF(values[__pyx_temp]);
}
__Pyx_AddTraceback("rugo.jsonl.read_jsonl._convert_strings_to_bytes_inplace", __pyx_clineno, __pyx_lineno, __pyx_filename);
__Pyx_RefNannyFinishContext();
return NULL;
__pyx_L4_argument_unpacking_done:;
__pyx_r = __pyx_pf_4rugo_5jsonl_10read_jsonl__convert_strings_to_bytes_inplace(__pyx_self, __pyx_v_obj);
int __pyx_lineno = 0;
const char *__pyx_filename = NULL;
int __pyx_clineno = 0;
/* function exit code */
for (Py_ssize_t __pyx_temp=0; __pyx_temp < (Py_ssize_t)(sizeof(values)/sizeof(values[0])); ++__pyx_temp) {
Py_XDECREF(values[__pyx_temp]);
}
__Pyx_RefNannyFinishContext();
return __pyx_r;
}
static PyObject *__pyx_pf_4rugo_5jsonl_10read_jsonl__convert_strings_to_bytes_inplace(PyObject *__pyx_self, PyObject *__pyx_v_obj) {
struct __pyx_obj_4rugo_5jsonl___pyx_scope_struct_1_read_jsonl *__pyx_cur_scope;
struct __pyx_obj_4rugo_5jsonl___pyx_scope_struct_1_read_jsonl *__pyx_outer_scope;
Py_ssize_t __pyx_v_idx;
PyObject *__pyx_v_v = NULL;
PyObject *__pyx_r = NULL;
__pyx_outer_scope = (struct __pyx_obj_4rugo_5jsonl___pyx_scope_struct_1_read_jsonl *) __Pyx_CyFunction_GetClosure(__pyx_self);
__pyx_cur_scope = __pyx_outer_scope;
/* … */
/* function exit code */
__pyx_r = Py_None; __Pyx_INCREF(Py_None);
goto __pyx_L0;
__pyx_L1_error:;
__Pyx_XDECREF(__pyx_t_5);
__Pyx_XDECREF(__pyx_t_6);
__Pyx_AddTraceback("rugo.jsonl.read_jsonl._convert_strings_to_bytes_inplace", __pyx_clineno, __pyx_lineno, __pyx_filename);
__pyx_r = NULL;
__pyx_L0:;
__Pyx_XDECREF(__pyx_v_v);
__Pyx_XGIVEREF(__pyx_r);
__Pyx_RefNannyFinishContext();
return __pyx_r;
}
/* … */
__pyx_t_7 = __Pyx_CyFunction_New(&__pyx_mdef_4rugo_5jsonl_10read_jsonl_1_convert_strings_to_bytes_inplace, 0, __pyx_mstate_global->__pyx_n_u_read_jsonl_locals__convert_strin, ((PyObject*)__pyx_cur_scope), __pyx_mstate_global->__pyx_n_u_rugo_jsonl, __pyx_mstate_global->__pyx_d, ((PyObject *)__pyx_mstate_global->__pyx_codeobj_tab[1])); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 407, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_7);
__Pyx_XGOTREF(__pyx_cur_scope->__pyx_v__convert_strings_to_bytes_inplace);
__Pyx_XDECREF_SET(__pyx_cur_scope->__pyx_v__convert_strings_to_bytes_inplace, __pyx_t_7);
__Pyx_GIVEREF(__pyx_t_7);
__pyx_t_7 = 0;
408: # Recursively convert str elements in lists to bytes when elem_type == 'bytes'
409: # obj is a Python object returned from the parser
+410: if isinstance(obj, list):
__pyx_t_1 = PyList_Check(__pyx_v_obj);
if (__pyx_t_1) {
/* … */
}
+411: for idx in range(len(obj)):
__pyx_t_2 = PyObject_Length(__pyx_v_obj); if (unlikely(__pyx_t_2 == ((Py_ssize_t)-1))) __PYX_ERR(0, 411, __pyx_L1_error) __pyx_t_3 = __pyx_t_2; for (__pyx_t_4 = 0; __pyx_t_4 < __pyx_t_3; __pyx_t_4+=1) { __pyx_v_idx = __pyx_t_4;
+412: v = obj[idx]
__pyx_t_5 = __Pyx_GetItemInt(__pyx_v_obj, __pyx_v_idx, Py_ssize_t, 1, PyLong_FromSsize_t, 0, 0, 0, 1); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 412, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_5); __Pyx_XDECREF_SET(__pyx_v_v, __pyx_t_5); __pyx_t_5 = 0;
+413: if isinstance(v, str):
__pyx_t_1 = PyUnicode_Check(__pyx_v_v);
if (__pyx_t_1) {
/* … */
goto __pyx_L6;
}
+414: obj[idx] = v.encode('utf-8')
__pyx_t_6 = __pyx_v_v;
__Pyx_INCREF(__pyx_t_6);
__pyx_t_7 = 0;
{
PyObject *__pyx_callargs[2] = {__pyx_t_6, __pyx_mstate_global->__pyx_kp_u_utf_8};
__pyx_t_5 = __Pyx_PyObject_FastCallMethod(__pyx_mstate_global->__pyx_n_u_encode, __pyx_callargs+__pyx_t_7, (2-__pyx_t_7) | (1*__Pyx_PY_VECTORCALL_ARGUMENTS_OFFSET));
__Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 414, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_5);
}
if (unlikely((__Pyx_SetItemInt(__pyx_v_obj, __pyx_v_idx, __pyx_t_5, Py_ssize_t, 1, PyLong_FromSsize_t, 0, 0, 0, 1) < 0))) __PYX_ERR(0, 414, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
+415: elif isinstance(v, list):
__pyx_t_1 = PyList_Check(__pyx_v_v);
if (__pyx_t_1) {
/* … */
}
__pyx_L6:;
}
+416: _convert_strings_to_bytes_inplace(v)
if (unlikely(!__pyx_cur_scope->__pyx_v__convert_strings_to_bytes_inplace)) { __Pyx_RaiseClosureNameError("_convert_strings_to_bytes_inplace"); __PYX_ERR(0, 416, __pyx_L1_error) }
__pyx_t_5 = __pyx_pf_4rugo_5jsonl_10read_jsonl__convert_strings_to_bytes_inplace(__pyx_cur_scope->__pyx_v__convert_strings_to_bytes_inplace, __pyx_v_v); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 416, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_5);
__Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
417: # leave dicts/bytes as-is
418:
+419: py_list = []
__pyx_t_7 = PyList_New(0); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 419, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_7); __Pyx_XDECREF_SET(__pyx_v_py_list, ((PyObject*)__pyx_t_7)); __pyx_t_7 = 0;
+420: for j in range(col.string_values.size()):
__pyx_t_12 = __pyx_v_col->string_values.size();
__pyx_t_13 = __pyx_t_12;
for (__pyx_t_22 = 0; __pyx_t_22 < __pyx_t_13; __pyx_t_22+=1) {
__pyx_v_j = __pyx_t_22;
+421: if col.null_mask[j]:
__pyx_t_1 = ((__pyx_v_col->null_mask[__pyx_v_j]) != 0);
if (__pyx_t_1) {
/* … */
}
+422: py_list.append(None)
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_list, Py_None); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 422, __pyx_L1_error)
+423: continue
goto __pyx_L33_continue;
424:
+425: raw = col.string_values[j]
__pyx_v_raw = (__pyx_v_col->string_values[__pyx_v_j]);
+426: if raw.size() == 0:
__pyx_t_1 = (__pyx_v_raw.size() == 0);
if (__pyx_t_1) {
/* … */
}
+427: py_list.append([])
__pyx_t_7 = PyList_New(0); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 427, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_7); __pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_list, __pyx_t_7); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 427, __pyx_L1_error) __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
+428: continue
goto __pyx_L33_continue;
429:
+430: if parse_arrays:
__pyx_t_1 = __Pyx_PyObject_IsTrue(__pyx_v_parse_arrays); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 430, __pyx_L1_error) if (__pyx_t_1) { /* … */ goto __pyx_L37; }
431: # Parse the JSON array into Python list
+432: o_ptr = ParseJsonSliceToPyObject(<const uint8_t*>raw.data(), raw.size(), True)
__pyx_v_o_ptr = ParseJsonSliceToPyObject(((uint8_t const *)__pyx_v_raw.data()), __pyx_v_raw.size(), 1);
+433: if o_ptr != NULL:
__pyx_t_1 = (__pyx_v_o_ptr != NULL);
if (__pyx_t_1) {
/* … */
goto __pyx_L38;
}
+434: o = <object>o_ptr
__pyx_t_7 = ((PyObject *)__pyx_v_o_ptr);
__Pyx_INCREF(__pyx_t_7);
__Pyx_DECREF_SET(__pyx_v_o, __pyx_t_7);
__pyx_t_7 = 0;
435: # If element type is bytes, or unspecified but the parsed
436: # array contains string elements (likely binary-as-JSON
437: # strings), convert those strings to bytes.
+438: if isinstance(o, list):
__pyx_t_1 = PyList_Check(__pyx_v_o);
if (__pyx_t_1) {
/* … */
}
+439: if elem_type == 'bytes':
__pyx_t_1 = (__Pyx_PyUnicode_Equals(__pyx_v_elem_type, __pyx_mstate_global->__pyx_n_u_bytes, Py_EQ)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 439, __pyx_L1_error) if (__pyx_t_1) { /* … */ goto __pyx_L40; }
+440: _convert_strings_to_bytes_inplace(o)
__pyx_t_7 = __pyx_pf_4rugo_5jsonl_10read_jsonl__convert_strings_to_bytes_inplace(__pyx_cur_scope->__pyx_v__convert_strings_to_bytes_inplace, __pyx_v_o); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 440, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_7); __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
+441: elif elem_type is None:
__pyx_t_1 = (__pyx_v_elem_type == Py_None);
if (__pyx_t_1) {
/* … */
}
__pyx_L40:;
442: # Heuristic: if at least one leaf element is str,
443: # convert all string leaves to bytes
+444: def _has_string_leaf(x):
/* Python wrapper */
static PyObject *__pyx_pw_4rugo_5jsonl_10read_jsonl_3_has_string_leaf(PyObject *__pyx_self,
#if CYTHON_METH_FASTCALL
PyObject *const *__pyx_args, Py_ssize_t __pyx_nargs, PyObject *__pyx_kwds
#else
PyObject *__pyx_args, PyObject *__pyx_kwds
#endif
); /*proto*/
static PyMethodDef __pyx_mdef_4rugo_5jsonl_10read_jsonl_3_has_string_leaf = {"_has_string_leaf", (PyCFunction)(void(*)(void))(__Pyx_PyCFunction_FastCallWithKeywords)__pyx_pw_4rugo_5jsonl_10read_jsonl_3_has_string_leaf, __Pyx_METH_FASTCALL|METH_KEYWORDS, 0};
static PyObject *__pyx_pw_4rugo_5jsonl_10read_jsonl_3_has_string_leaf(PyObject *__pyx_self,
#if CYTHON_METH_FASTCALL
PyObject *const *__pyx_args, Py_ssize_t __pyx_nargs, PyObject *__pyx_kwds
#else
PyObject *__pyx_args, PyObject *__pyx_kwds
#endif
) {
PyObject *__pyx_v_x = 0;
#if !CYTHON_METH_FASTCALL
CYTHON_UNUSED Py_ssize_t __pyx_nargs;
#endif
CYTHON_UNUSED PyObject *const *__pyx_kwvalues;
PyObject *__pyx_r = 0;
__Pyx_RefNannyDeclarations
__Pyx_RefNannySetupContext("_has_string_leaf (wrapper)", 0);
#if !CYTHON_METH_FASTCALL
#if CYTHON_ASSUME_SAFE_SIZE
__pyx_nargs = PyTuple_GET_SIZE(__pyx_args);
#else
__pyx_nargs = PyTuple_Size(__pyx_args); if (unlikely(__pyx_nargs < 0)) return NULL;
#endif
#endif
__pyx_kwvalues = __Pyx_KwValues_FASTCALL(__pyx_args, __pyx_nargs);
{
PyObject ** const __pyx_pyargnames[] = {&__pyx_mstate_global->__pyx_n_u_x,0};
PyObject* values[1] = {0};
const Py_ssize_t __pyx_kwds_len = (__pyx_kwds) ? __Pyx_NumKwargs_FASTCALL(__pyx_kwds) : 0;
if (unlikely(__pyx_kwds_len) < 0) __PYX_ERR(0, 444, __pyx_L3_error)
if (__pyx_kwds_len > 0) {
switch (__pyx_nargs) {
case 1:
values[0] = __Pyx_ArgRef_FASTCALL(__pyx_args, 0);
if (!CYTHON_ASSUME_SAFE_MACROS && unlikely(!values[0])) __PYX_ERR(0, 444, __pyx_L3_error)
CYTHON_FALLTHROUGH;
case 0: break;
default: goto __pyx_L5_argtuple_error;
}
const Py_ssize_t kwd_pos_args = __pyx_nargs;
if (__Pyx_ParseKeywords(__pyx_kwds, __pyx_kwvalues, __pyx_pyargnames, 0, values, kwd_pos_args, __pyx_kwds_len, "_has_string_leaf", 0) < (0)) __PYX_ERR(0, 444, __pyx_L3_error)
for (Py_ssize_t i = __pyx_nargs; i < 1; i++) {
if (unlikely(!values[i])) { __Pyx_RaiseArgtupleInvalid("_has_string_leaf", 1, 1, 1, i); __PYX_ERR(0, 444, __pyx_L3_error) }
}
} else if (unlikely(__pyx_nargs != 1)) {
goto __pyx_L5_argtuple_error;
} else {
values[0] = __Pyx_ArgRef_FASTCALL(__pyx_args, 0);
if (!CYTHON_ASSUME_SAFE_MACROS && unlikely(!values[0])) __PYX_ERR(0, 444, __pyx_L3_error)
}
__pyx_v_x = values[0];
}
goto __pyx_L6_skip;
__pyx_L5_argtuple_error:;
__Pyx_RaiseArgtupleInvalid("_has_string_leaf", 1, 1, 1, __pyx_nargs); __PYX_ERR(0, 444, __pyx_L3_error)
__pyx_L6_skip:;
goto __pyx_L4_argument_unpacking_done;
__pyx_L3_error:;
for (Py_ssize_t __pyx_temp=0; __pyx_temp < (Py_ssize_t)(sizeof(values)/sizeof(values[0])); ++__pyx_temp) {
Py_XDECREF(values[__pyx_temp]);
}
__Pyx_AddTraceback("rugo.jsonl.read_jsonl._has_string_leaf", __pyx_clineno, __pyx_lineno, __pyx_filename);
__Pyx_RefNannyFinishContext();
return NULL;
__pyx_L4_argument_unpacking_done:;
__pyx_r = __pyx_pf_4rugo_5jsonl_10read_jsonl_2_has_string_leaf(__pyx_self, __pyx_v_x);
int __pyx_lineno = 0;
const char *__pyx_filename = NULL;
int __pyx_clineno = 0;
/* function exit code */
for (Py_ssize_t __pyx_temp=0; __pyx_temp < (Py_ssize_t)(sizeof(values)/sizeof(values[0])); ++__pyx_temp) {
Py_XDECREF(values[__pyx_temp]);
}
__Pyx_RefNannyFinishContext();
return __pyx_r;
}
static PyObject *__pyx_pf_4rugo_5jsonl_10read_jsonl_2_has_string_leaf(PyObject *__pyx_self, PyObject *__pyx_v_x) {
struct __pyx_obj_4rugo_5jsonl___pyx_scope_struct_1_read_jsonl *__pyx_cur_scope;
struct __pyx_obj_4rugo_5jsonl___pyx_scope_struct_1_read_jsonl *__pyx_outer_scope;
PyObject *__pyx_v_y = NULL;
PyObject *__pyx_r = NULL;
__pyx_outer_scope = (struct __pyx_obj_4rugo_5jsonl___pyx_scope_struct_1_read_jsonl *) __Pyx_CyFunction_GetClosure(__pyx_self);
__pyx_cur_scope = __pyx_outer_scope;
/* … */
/* function exit code */
__pyx_L1_error:;
__Pyx_XDECREF(__pyx_t_2);
__Pyx_XDECREF(__pyx_t_5);
__Pyx_AddTraceback("rugo.jsonl.read_jsonl._has_string_leaf", __pyx_clineno, __pyx_lineno, __pyx_filename);
__pyx_r = NULL;
__pyx_L0:;
__Pyx_XDECREF(__pyx_v_y);
__Pyx_XGIVEREF(__pyx_r);
__Pyx_RefNannyFinishContext();
return __pyx_r;
}
/* … */
__pyx_t_7 = __Pyx_CyFunction_New(&__pyx_mdef_4rugo_5jsonl_10read_jsonl_3_has_string_leaf, 0, __pyx_mstate_global->__pyx_n_u_read_jsonl_locals__has_string_le, ((PyObject*)__pyx_cur_scope), __pyx_mstate_global->__pyx_n_u_rugo_jsonl, __pyx_mstate_global->__pyx_d, ((PyObject *)__pyx_mstate_global->__pyx_codeobj_tab[2])); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 444, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_7);
__Pyx_XGOTREF(__pyx_cur_scope->__pyx_v__has_string_leaf);
__Pyx_XDECREF_SET(__pyx_cur_scope->__pyx_v__has_string_leaf, __pyx_t_7);
__Pyx_GIVEREF(__pyx_t_7);
__pyx_t_7 = 0;
+445: if isinstance(x, list):
__pyx_t_1 = PyList_Check(__pyx_v_x);
if (__pyx_t_1) {
/* … */
}
+446: for y in x:
if (likely(PyList_CheckExact(__pyx_v_x)) || PyTuple_CheckExact(__pyx_v_x)) { __pyx_t_2 = __pyx_v_x; __Pyx_INCREF(__pyx_t_2); __pyx_t_3 = 0; __pyx_t_4 = NULL; } else { __pyx_t_3 = -1; __pyx_t_2 = PyObject_GetIter(__pyx_v_x); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 446, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_2); __pyx_t_4 = (CYTHON_COMPILING_IN_LIMITED_API) ? PyIter_Next : __Pyx_PyObject_GetIterNextFunc(__pyx_t_2); if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 446, __pyx_L1_error) } for (;;) { if (likely(!__pyx_t_4)) { if (likely(PyList_CheckExact(__pyx_t_2))) { { Py_ssize_t __pyx_temp = __Pyx_PyList_GET_SIZE(__pyx_t_2); #if !CYTHON_ASSUME_SAFE_SIZE if (unlikely((__pyx_temp < 0))) __PYX_ERR(0, 446, __pyx_L1_error) #endif if (__pyx_t_3 >= __pyx_temp) break; } __pyx_t_5 = __Pyx_PyList_GetItemRef(__pyx_t_2, __pyx_t_3); ++__pyx_t_3; } else { { Py_ssize_t __pyx_temp = __Pyx_PyTuple_GET_SIZE(__pyx_t_2); #if !CYTHON_ASSUME_SAFE_SIZE if (unlikely((__pyx_temp < 0))) __PYX_ERR(0, 446, __pyx_L1_error) #endif if (__pyx_t_3 >= __pyx_temp) break; } #if CYTHON_ASSUME_SAFE_MACROS && !CYTHON_AVOID_BORROWED_REFS __pyx_t_5 = __Pyx_NewRef(PyTuple_GET_ITEM(__pyx_t_2, __pyx_t_3)); #else __pyx_t_5 = __Pyx_PySequence_ITEM(__pyx_t_2, __pyx_t_3); #endif ++__pyx_t_3; } if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 446, __pyx_L1_error) } else { __pyx_t_5 = __pyx_t_4(__pyx_t_2); if (unlikely(!__pyx_t_5)) { PyObject* exc_type = PyErr_Occurred(); if (exc_type) { if (unlikely(!__Pyx_PyErr_GivenExceptionMatches(exc_type, PyExc_StopIteration))) __PYX_ERR(0, 446, __pyx_L1_error) PyErr_Clear(); } break; } } __Pyx_GOTREF(__pyx_t_5); __Pyx_XDECREF_SET(__pyx_v_y, __pyx_t_5); __pyx_t_5 = 0; /* … */ } __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
+447: if _has_string_leaf(y):
if (unlikely(!__pyx_cur_scope->__pyx_v__has_string_leaf)) { __Pyx_RaiseClosureNameError("_has_string_leaf"); __PYX_ERR(0, 447, __pyx_L1_error) }
__pyx_t_5 = __Pyx_PyObject_CallOneArg(__pyx_cur_scope->__pyx_v__has_string_leaf, __pyx_v_y); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 447, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_5);
__pyx_t_1 = __Pyx_PyObject_IsTrue(__pyx_t_5); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 447, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
if (__pyx_t_1) {
/* … */
}
+448: return True
__Pyx_XDECREF(__pyx_r); __Pyx_INCREF(Py_True); __pyx_r = Py_True; __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; goto __pyx_L0;
+449: return False
__Pyx_XDECREF(__pyx_r); __Pyx_INCREF(Py_False); __pyx_r = Py_False; goto __pyx_L0;
+450: return isinstance(x, str)
__Pyx_XDECREF(__pyx_r); __pyx_t_1 = PyUnicode_Check(__pyx_v_x); __pyx_t_2 = __Pyx_PyBool_FromLong(__pyx_t_1); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 450, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_2); __pyx_r = __pyx_t_2; __pyx_t_2 = 0; goto __pyx_L0;
+451: if _has_string_leaf(o):
__pyx_t_7 = __Pyx_PyObject_CallOneArg(__pyx_cur_scope->__pyx_v__has_string_leaf, __pyx_v_o); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 451, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_7); __pyx_t_1 = __Pyx_PyObject_IsTrue(__pyx_t_7); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 451, __pyx_L1_error) __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0; if (__pyx_t_1) { /* … */ }
+452: _convert_strings_to_bytes_inplace(o)
__pyx_t_7 = __pyx_pf_4rugo_5jsonl_10read_jsonl__convert_strings_to_bytes_inplace(__pyx_cur_scope->__pyx_v__convert_strings_to_bytes_inplace, __pyx_v_o); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 452, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_7); __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
+453: py_list.append(o)
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_list, __pyx_v_o); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 453, __pyx_L1_error)
454: else:
+455: if PyErr_Occurred():
/*else*/ {
__pyx_t_1 = (PyErr_Occurred() != 0);
if (__pyx_t_1) {
/* … */
}
+456: PyErr_Clear()
PyErr_Clear();
+457: try:
{
/*try:*/ {
/* … */
}
__Pyx_XDECREF(__pyx_t_24); __pyx_t_24 = 0;
__Pyx_XDECREF(__pyx_t_25); __pyx_t_25 = 0;
__Pyx_XDECREF(__pyx_t_26); __pyx_t_26 = 0;
goto __pyx_L50_try_end;
__pyx_L43_error:;
__Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
__Pyx_XDECREF(__pyx_t_27); __pyx_t_27 = 0;
__Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
__Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
/* … */
__pyx_L45_except_error:;
__Pyx_XGIVEREF(__pyx_t_24);
__Pyx_XGIVEREF(__pyx_t_25);
__Pyx_XGIVEREF(__pyx_t_26);
__Pyx_ExceptionReset(__pyx_t_24, __pyx_t_25, __pyx_t_26);
goto __pyx_L1_error;
__pyx_L44_exception_handled:;
__Pyx_XGIVEREF(__pyx_t_24);
__Pyx_XGIVEREF(__pyx_t_25);
__Pyx_XGIVEREF(__pyx_t_26);
__Pyx_ExceptionReset(__pyx_t_24, __pyx_t_25, __pyx_t_26);
__pyx_L50_try_end:;
}
}
__pyx_L38:;
+458: parsed = _parse_array_from_bytes(raw)
__pyx_t_2 = NULL;
__Pyx_GetModuleGlobalName(__pyx_t_6, __pyx_mstate_global->__pyx_n_u_parse_array_from_bytes); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 458, __pyx_L43_error)
__Pyx_GOTREF(__pyx_t_6);
__pyx_t_27 = __pyx_convert_PyBytes_string_to_py_6libcpp_6string_std__in_string(__pyx_v_raw); if (unlikely(!__pyx_t_27)) __PYX_ERR(0, 458, __pyx_L43_error)
__Pyx_GOTREF(__pyx_t_27);
__pyx_t_23 = 1;
#if CYTHON_UNPACK_METHODS
if (unlikely(PyMethod_Check(__pyx_t_6))) {
__pyx_t_2 = PyMethod_GET_SELF(__pyx_t_6);
assert(__pyx_t_2);
PyObject* __pyx__function = PyMethod_GET_FUNCTION(__pyx_t_6);
__Pyx_INCREF(__pyx_t_2);
__Pyx_INCREF(__pyx__function);
__Pyx_DECREF_SET(__pyx_t_6, __pyx__function);
__pyx_t_23 = 0;
}
#endif
{
PyObject *__pyx_callargs[2] = {__pyx_t_2, __pyx_t_27};
__pyx_t_7 = __Pyx_PyObject_FastCall(__pyx_t_6, __pyx_callargs+__pyx_t_23, (2-__pyx_t_23) | (__pyx_t_23*__Pyx_PY_VECTORCALL_ARGUMENTS_OFFSET));
__Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
__Pyx_DECREF(__pyx_t_27); __pyx_t_27 = 0;
__Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 458, __pyx_L43_error)
__Pyx_GOTREF(__pyx_t_7);
}
__Pyx_XDECREF_SET(__pyx_v_parsed, __pyx_t_7);
__pyx_t_7 = 0;
+459: if isinstance(parsed, list):
__pyx_t_1 = PyList_Check(__pyx_v_parsed);
if (__pyx_t_1) {
/* … */
}
+460: if elem_type == 'bytes':
__pyx_t_1 = (__Pyx_PyUnicode_Equals(__pyx_v_elem_type, __pyx_mstate_global->__pyx_n_u_bytes, Py_EQ)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 460, __pyx_L43_error) if (__pyx_t_1) { /* … */ goto __pyx_L52; }
+461: _convert_strings_to_bytes_inplace(parsed)
__pyx_t_7 = __pyx_pf_4rugo_5jsonl_10read_jsonl__convert_strings_to_bytes_inplace(__pyx_cur_scope->__pyx_v__convert_strings_to_bytes_inplace, __pyx_v_parsed); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 461, __pyx_L43_error) __Pyx_GOTREF(__pyx_t_7); __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
+462: elif elem_type is None:
__pyx_t_1 = (__pyx_v_elem_type == Py_None);
if (__pyx_t_1) {
/* … */
}
__pyx_L52:;
+463: def _has_string_leaf(x):
/* Python wrapper */
static PyObject *__pyx_pw_4rugo_5jsonl_10read_jsonl_5_has_string_leaf(PyObject *__pyx_self,
#if CYTHON_METH_FASTCALL
PyObject *const *__pyx_args, Py_ssize_t __pyx_nargs, PyObject *__pyx_kwds
#else
PyObject *__pyx_args, PyObject *__pyx_kwds
#endif
); /*proto*/
static PyMethodDef __pyx_mdef_4rugo_5jsonl_10read_jsonl_5_has_string_leaf = {"_has_string_leaf", (PyCFunction)(void(*)(void))(__Pyx_PyCFunction_FastCallWithKeywords)__pyx_pw_4rugo_5jsonl_10read_jsonl_5_has_string_leaf, __Pyx_METH_FASTCALL|METH_KEYWORDS, 0};
static PyObject *__pyx_pw_4rugo_5jsonl_10read_jsonl_5_has_string_leaf(PyObject *__pyx_self,
#if CYTHON_METH_FASTCALL
PyObject *const *__pyx_args, Py_ssize_t __pyx_nargs, PyObject *__pyx_kwds
#else
PyObject *__pyx_args, PyObject *__pyx_kwds
#endif
) {
PyObject *__pyx_v_x = 0;
#if !CYTHON_METH_FASTCALL
CYTHON_UNUSED Py_ssize_t __pyx_nargs;
#endif
CYTHON_UNUSED PyObject *const *__pyx_kwvalues;
PyObject *__pyx_r = 0;
__Pyx_RefNannyDeclarations
__Pyx_RefNannySetupContext("_has_string_leaf (wrapper)", 0);
#if !CYTHON_METH_FASTCALL
#if CYTHON_ASSUME_SAFE_SIZE
__pyx_nargs = PyTuple_GET_SIZE(__pyx_args);
#else
__pyx_nargs = PyTuple_Size(__pyx_args); if (unlikely(__pyx_nargs < 0)) return NULL;
#endif
#endif
__pyx_kwvalues = __Pyx_KwValues_FASTCALL(__pyx_args, __pyx_nargs);
{
PyObject ** const __pyx_pyargnames[] = {&__pyx_mstate_global->__pyx_n_u_x,0};
PyObject* values[1] = {0};
const Py_ssize_t __pyx_kwds_len = (__pyx_kwds) ? __Pyx_NumKwargs_FASTCALL(__pyx_kwds) : 0;
if (unlikely(__pyx_kwds_len) < 0) __PYX_ERR(0, 463, __pyx_L3_error)
if (__pyx_kwds_len > 0) {
switch (__pyx_nargs) {
case 1:
values[0] = __Pyx_ArgRef_FASTCALL(__pyx_args, 0);
if (!CYTHON_ASSUME_SAFE_MACROS && unlikely(!values[0])) __PYX_ERR(0, 463, __pyx_L3_error)
CYTHON_FALLTHROUGH;
case 0: break;
default: goto __pyx_L5_argtuple_error;
}
const Py_ssize_t kwd_pos_args = __pyx_nargs;
if (__Pyx_ParseKeywords(__pyx_kwds, __pyx_kwvalues, __pyx_pyargnames, 0, values, kwd_pos_args, __pyx_kwds_len, "_has_string_leaf", 0) < (0)) __PYX_ERR(0, 463, __pyx_L3_error)
for (Py_ssize_t i = __pyx_nargs; i < 1; i++) {
if (unlikely(!values[i])) { __Pyx_RaiseArgtupleInvalid("_has_string_leaf", 1, 1, 1, i); __PYX_ERR(0, 463, __pyx_L3_error) }
}
} else if (unlikely(__pyx_nargs != 1)) {
goto __pyx_L5_argtuple_error;
} else {
values[0] = __Pyx_ArgRef_FASTCALL(__pyx_args, 0);
if (!CYTHON_ASSUME_SAFE_MACROS && unlikely(!values[0])) __PYX_ERR(0, 463, __pyx_L3_error)
}
__pyx_v_x = values[0];
}
goto __pyx_L6_skip;
__pyx_L5_argtuple_error:;
__Pyx_RaiseArgtupleInvalid("_has_string_leaf", 1, 1, 1, __pyx_nargs); __PYX_ERR(0, 463, __pyx_L3_error)
__pyx_L6_skip:;
goto __pyx_L4_argument_unpacking_done;
__pyx_L3_error:;
for (Py_ssize_t __pyx_temp=0; __pyx_temp < (Py_ssize_t)(sizeof(values)/sizeof(values[0])); ++__pyx_temp) {
Py_XDECREF(values[__pyx_temp]);
}
__Pyx_AddTraceback("rugo.jsonl.read_jsonl._has_string_leaf", __pyx_clineno, __pyx_lineno, __pyx_filename);
__Pyx_RefNannyFinishContext();
return NULL;
__pyx_L4_argument_unpacking_done:;
__pyx_r = __pyx_pf_4rugo_5jsonl_10read_jsonl_4_has_string_leaf(__pyx_self, __pyx_v_x);
int __pyx_lineno = 0;
const char *__pyx_filename = NULL;
int __pyx_clineno = 0;
/* function exit code */
for (Py_ssize_t __pyx_temp=0; __pyx_temp < (Py_ssize_t)(sizeof(values)/sizeof(values[0])); ++__pyx_temp) {
Py_XDECREF(values[__pyx_temp]);
}
__Pyx_RefNannyFinishContext();
return __pyx_r;
}
static PyObject *__pyx_pf_4rugo_5jsonl_10read_jsonl_4_has_string_leaf(PyObject *__pyx_self, PyObject *__pyx_v_x) {
struct __pyx_obj_4rugo_5jsonl___pyx_scope_struct_1_read_jsonl *__pyx_cur_scope;
struct __pyx_obj_4rugo_5jsonl___pyx_scope_struct_1_read_jsonl *__pyx_outer_scope;
PyObject *__pyx_v_y = NULL;
PyObject *__pyx_r = NULL;
__pyx_outer_scope = (struct __pyx_obj_4rugo_5jsonl___pyx_scope_struct_1_read_jsonl *) __Pyx_CyFunction_GetClosure(__pyx_self);
__pyx_cur_scope = __pyx_outer_scope;
/* … */
/* function exit code */
__pyx_L1_error:;
__Pyx_XDECREF(__pyx_t_2);
__Pyx_XDECREF(__pyx_t_5);
__Pyx_AddTraceback("rugo.jsonl.read_jsonl._has_string_leaf", __pyx_clineno, __pyx_lineno, __pyx_filename);
__pyx_r = NULL;
__pyx_L0:;
__Pyx_XDECREF(__pyx_v_y);
__Pyx_XGIVEREF(__pyx_r);
__Pyx_RefNannyFinishContext();
return __pyx_r;
}
/* … */
__pyx_t_7 = __Pyx_CyFunction_New(&__pyx_mdef_4rugo_5jsonl_10read_jsonl_5_has_string_leaf, 0, __pyx_mstate_global->__pyx_n_u_read_jsonl_locals__has_string_le, ((PyObject*)__pyx_cur_scope), __pyx_mstate_global->__pyx_n_u_rugo_jsonl, __pyx_mstate_global->__pyx_d, ((PyObject *)__pyx_mstate_global->__pyx_codeobj_tab[3])); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 463, __pyx_L43_error)
__Pyx_GOTREF(__pyx_t_7);
__Pyx_XGOTREF(__pyx_cur_scope->__pyx_v__has_string_leaf);
__Pyx_XDECREF_SET(__pyx_cur_scope->__pyx_v__has_string_leaf, __pyx_t_7);
__Pyx_GIVEREF(__pyx_t_7);
__pyx_t_7 = 0;
+464: if isinstance(x, list):
__pyx_t_1 = PyList_Check(__pyx_v_x);
if (__pyx_t_1) {
/* … */
}
+465: for y in x:
if (likely(PyList_CheckExact(__pyx_v_x)) || PyTuple_CheckExact(__pyx_v_x)) { __pyx_t_2 = __pyx_v_x; __Pyx_INCREF(__pyx_t_2); __pyx_t_3 = 0; __pyx_t_4 = NULL; } else { __pyx_t_3 = -1; __pyx_t_2 = PyObject_GetIter(__pyx_v_x); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 465, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_2); __pyx_t_4 = (CYTHON_COMPILING_IN_LIMITED_API) ? PyIter_Next : __Pyx_PyObject_GetIterNextFunc(__pyx_t_2); if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 465, __pyx_L1_error) } for (;;) { if (likely(!__pyx_t_4)) { if (likely(PyList_CheckExact(__pyx_t_2))) { { Py_ssize_t __pyx_temp = __Pyx_PyList_GET_SIZE(__pyx_t_2); #if !CYTHON_ASSUME_SAFE_SIZE if (unlikely((__pyx_temp < 0))) __PYX_ERR(0, 465, __pyx_L1_error) #endif if (__pyx_t_3 >= __pyx_temp) break; } __pyx_t_5 = __Pyx_PyList_GetItemRef(__pyx_t_2, __pyx_t_3); ++__pyx_t_3; } else { { Py_ssize_t __pyx_temp = __Pyx_PyTuple_GET_SIZE(__pyx_t_2); #if !CYTHON_ASSUME_SAFE_SIZE if (unlikely((__pyx_temp < 0))) __PYX_ERR(0, 465, __pyx_L1_error) #endif if (__pyx_t_3 >= __pyx_temp) break; } #if CYTHON_ASSUME_SAFE_MACROS && !CYTHON_AVOID_BORROWED_REFS __pyx_t_5 = __Pyx_NewRef(PyTuple_GET_ITEM(__pyx_t_2, __pyx_t_3)); #else __pyx_t_5 = __Pyx_PySequence_ITEM(__pyx_t_2, __pyx_t_3); #endif ++__pyx_t_3; } if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 465, __pyx_L1_error) } else { __pyx_t_5 = __pyx_t_4(__pyx_t_2); if (unlikely(!__pyx_t_5)) { PyObject* exc_type = PyErr_Occurred(); if (exc_type) { if (unlikely(!__Pyx_PyErr_GivenExceptionMatches(exc_type, PyExc_StopIteration))) __PYX_ERR(0, 465, __pyx_L1_error) PyErr_Clear(); } break; } } __Pyx_GOTREF(__pyx_t_5); __Pyx_XDECREF_SET(__pyx_v_y, __pyx_t_5); __pyx_t_5 = 0; /* … */ } __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
+466: if _has_string_leaf(y):
if (unlikely(!__pyx_cur_scope->__pyx_v__has_string_leaf)) { __Pyx_RaiseClosureNameError("_has_string_leaf"); __PYX_ERR(0, 466, __pyx_L1_error) }
__pyx_t_5 = __Pyx_PyObject_CallOneArg(__pyx_cur_scope->__pyx_v__has_string_leaf, __pyx_v_y); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 466, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_5);
__pyx_t_1 = __Pyx_PyObject_IsTrue(__pyx_t_5); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 466, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
if (__pyx_t_1) {
/* … */
}
+467: return True
__Pyx_XDECREF(__pyx_r); __Pyx_INCREF(Py_True); __pyx_r = Py_True; __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; goto __pyx_L0;
+468: return False
__Pyx_XDECREF(__pyx_r); __Pyx_INCREF(Py_False); __pyx_r = Py_False; goto __pyx_L0;
+469: return isinstance(x, str)
__Pyx_XDECREF(__pyx_r); __pyx_t_1 = PyUnicode_Check(__pyx_v_x); __pyx_t_2 = __Pyx_PyBool_FromLong(__pyx_t_1); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 469, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_2); __pyx_r = __pyx_t_2; __pyx_t_2 = 0; goto __pyx_L0;
+470: if _has_string_leaf(parsed):
__pyx_t_7 = __Pyx_PyObject_CallOneArg(__pyx_cur_scope->__pyx_v__has_string_leaf, __pyx_v_parsed); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 470, __pyx_L43_error) __Pyx_GOTREF(__pyx_t_7); __pyx_t_1 = __Pyx_PyObject_IsTrue(__pyx_t_7); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 470, __pyx_L43_error) __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0; if (__pyx_t_1) { /* … */ }
+471: _convert_strings_to_bytes_inplace(parsed)
__pyx_t_7 = __pyx_pf_4rugo_5jsonl_10read_jsonl__convert_strings_to_bytes_inplace(__pyx_cur_scope->__pyx_v__convert_strings_to_bytes_inplace, __pyx_v_parsed); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 471, __pyx_L43_error) __Pyx_GOTREF(__pyx_t_7); __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
+472: py_list.append(parsed)
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_list, __pyx_v_parsed); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 472, __pyx_L43_error)
+473: except Exception:
__pyx_t_5 = __Pyx_PyErr_ExceptionMatches(((PyObject *)(((PyTypeObject*)PyExc_Exception)))); if (__pyx_t_5) { __Pyx_AddTraceback("rugo.jsonl.read_jsonl", __pyx_clineno, __pyx_lineno, __pyx_filename); if (__Pyx_GetException(&__pyx_t_7, &__pyx_t_6, &__pyx_t_27) < 0) __PYX_ERR(0, 473, __pyx_L45_except_error) __Pyx_XGOTREF(__pyx_t_7); __Pyx_XGOTREF(__pyx_t_6); __Pyx_XGOTREF(__pyx_t_27);
474: # Fallback to raw string
+475: py_list.append(raw.decode('utf-8'))
__pyx_t_2 = __Pyx_decode_cpp_string(__pyx_v_raw, 0, PY_SSIZE_T_MAX, NULL, NULL, PyUnicode_DecodeUTF8); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 475, __pyx_L45_except_error) __Pyx_GOTREF(__pyx_t_2); __pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_list, __pyx_t_2); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 475, __pyx_L45_except_error) __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0; __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0; __Pyx_XDECREF(__pyx_t_27); __pyx_t_27 = 0; goto __pyx_L44_exception_handled; } goto __pyx_L45_except_error;
476: else:
477: # Return as string without parsing
+478: py_list.append(raw.decode('utf-8'))
/*else*/ {
__pyx_t_27 = __Pyx_decode_cpp_string(__pyx_v_raw, 0, PY_SSIZE_T_MAX, NULL, NULL, PyUnicode_DecodeUTF8); if (unlikely(!__pyx_t_27)) __PYX_ERR(0, 478, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_27);
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_list, __pyx_t_27); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 478, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_27); __pyx_t_27 = 0;
}
__pyx_L37:;
__pyx_L33_continue:;
}
+479: py_columns.append(py_list)
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_columns, __pyx_v_py_list); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 479, __pyx_L1_error)
+480: elif col_type == 'object':
__pyx_t_1 = (__Pyx_PyUnicode_Equals(__pyx_v_col_type, __pyx_mstate_global->__pyx_n_u_object, Py_EQ)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 480, __pyx_L1_error) if (__pyx_t_1) { /* … */ goto __pyx_L16; }
481: # Object columns: return as JSONB (bytes), may contain objects, arrays, or mixed
482: # Check each value and handle appropriately
+483: py_list = []
__pyx_t_27 = PyList_New(0); if (unlikely(!__pyx_t_27)) __PYX_ERR(0, 483, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_27); __Pyx_XDECREF_SET(__pyx_v_py_list, ((PyObject*)__pyx_t_27)); __pyx_t_27 = 0;
+484: for j in range(col.string_values.size()):
__pyx_t_12 = __pyx_v_col->string_values.size();
__pyx_t_13 = __pyx_t_12;
for (__pyx_t_22 = 0; __pyx_t_22 < __pyx_t_13; __pyx_t_22+=1) {
__pyx_v_j = __pyx_t_22;
+485: if col.null_mask[j]:
__pyx_t_1 = ((__pyx_v_col->null_mask[__pyx_v_j]) != 0);
if (__pyx_t_1) {
/* … */
}
+486: py_list.append(None)
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_list, Py_None); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 486, __pyx_L1_error)
+487: continue
goto __pyx_L56_continue;
488:
+489: raw = col.string_values[j]
__pyx_v_raw = (__pyx_v_col->string_values[__pyx_v_j]);
+490: if raw.size() == 0:
__pyx_t_1 = (__pyx_v_raw.size() == 0);
if (__pyx_t_1) {
/* … */
}
+491: py_list.append(b'{}')
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_list, __pyx_mstate_global->__pyx_kp_b__12); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 491, __pyx_L1_error)
+492: continue
goto __pyx_L56_continue;
493:
494: # Check what type of JSON value this is
+495: first = raw[0]
__pyx_v_first = (__pyx_v_raw[0]);
496:
+497: if first == '[':
switch (__pyx_v_first) {
case '[':
/* … */
break;
case '{':
498: # This is an array in a mixed column
+499: if parse_arrays:
__pyx_t_1 = __Pyx_PyObject_IsTrue(__pyx_v_parse_arrays); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 499, __pyx_L1_error) if (__pyx_t_1) { /* … */ goto __pyx_L60; }
500: # Parse as array into Python list
+501: o_ptr = ParseJsonSliceToPyObject(<const uint8_t*>raw.data(), raw.size(), True)
__pyx_v_o_ptr = ParseJsonSliceToPyObject(((uint8_t const *)__pyx_v_raw.data()), __pyx_v_raw.size(), 1);
+502: if o_ptr != NULL:
__pyx_t_1 = (__pyx_v_o_ptr != NULL);
if (__pyx_t_1) {
/* … */
goto __pyx_L61;
}
+503: o = <object>o_ptr
__pyx_t_27 = ((PyObject *)__pyx_v_o_ptr);
__Pyx_INCREF(__pyx_t_27);
__Pyx_DECREF_SET(__pyx_v_o, __pyx_t_27);
__pyx_t_27 = 0;
+504: py_list.append(o)
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_list, __pyx_v_o); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 504, __pyx_L1_error)
505: else:
+506: if PyErr_Occurred():
/*else*/ {
__pyx_t_1 = (PyErr_Occurred() != 0);
if (__pyx_t_1) {
/* … */
}
+507: PyErr_Clear()
PyErr_Clear();
+508: try:
{
/*try:*/ {
/* … */
}
__Pyx_XDECREF(__pyx_t_26); __pyx_t_26 = 0;
__Pyx_XDECREF(__pyx_t_25); __pyx_t_25 = 0;
__Pyx_XDECREF(__pyx_t_24); __pyx_t_24 = 0;
goto __pyx_L70_try_end;
__pyx_L63_error:;
__Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
__Pyx_XDECREF(__pyx_t_27); __pyx_t_27 = 0;
__Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
__Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
/* … */
__pyx_L65_except_error:;
__Pyx_XGIVEREF(__pyx_t_26);
__Pyx_XGIVEREF(__pyx_t_25);
__Pyx_XGIVEREF(__pyx_t_24);
__Pyx_ExceptionReset(__pyx_t_26, __pyx_t_25, __pyx_t_24);
goto __pyx_L1_error;
__pyx_L64_exception_handled:;
__Pyx_XGIVEREF(__pyx_t_26);
__Pyx_XGIVEREF(__pyx_t_25);
__Pyx_XGIVEREF(__pyx_t_24);
__Pyx_ExceptionReset(__pyx_t_26, __pyx_t_25, __pyx_t_24);
__pyx_L70_try_end:;
}
}
__pyx_L61:;
+509: parsed = _parse_array_from_bytes(raw)
__pyx_t_6 = NULL;
__Pyx_GetModuleGlobalName(__pyx_t_7, __pyx_mstate_global->__pyx_n_u_parse_array_from_bytes); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 509, __pyx_L63_error)
__Pyx_GOTREF(__pyx_t_7);
__pyx_t_2 = __pyx_convert_PyBytes_string_to_py_6libcpp_6string_std__in_string(__pyx_v_raw); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 509, __pyx_L63_error)
__Pyx_GOTREF(__pyx_t_2);
__pyx_t_23 = 1;
#if CYTHON_UNPACK_METHODS
if (unlikely(PyMethod_Check(__pyx_t_7))) {
__pyx_t_6 = PyMethod_GET_SELF(__pyx_t_7);
assert(__pyx_t_6);
PyObject* __pyx__function = PyMethod_GET_FUNCTION(__pyx_t_7);
__Pyx_INCREF(__pyx_t_6);
__Pyx_INCREF(__pyx__function);
__Pyx_DECREF_SET(__pyx_t_7, __pyx__function);
__pyx_t_23 = 0;
}
#endif
{
PyObject *__pyx_callargs[2] = {__pyx_t_6, __pyx_t_2};
__pyx_t_27 = __Pyx_PyObject_FastCall(__pyx_t_7, __pyx_callargs+__pyx_t_23, (2-__pyx_t_23) | (__pyx_t_23*__Pyx_PY_VECTORCALL_ARGUMENTS_OFFSET));
__Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
__Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
__Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
if (unlikely(!__pyx_t_27)) __PYX_ERR(0, 509, __pyx_L63_error)
__Pyx_GOTREF(__pyx_t_27);
}
__Pyx_XDECREF_SET(__pyx_v_parsed, __pyx_t_27);
__pyx_t_27 = 0;
+510: py_list.append(parsed)
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_list, __pyx_v_parsed); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 510, __pyx_L63_error)
+511: except Exception:
__pyx_t_5 = __Pyx_PyErr_ExceptionMatches(((PyObject *)(((PyTypeObject*)PyExc_Exception)))); if (__pyx_t_5) { __Pyx_AddTraceback("rugo.jsonl.read_jsonl", __pyx_clineno, __pyx_lineno, __pyx_filename); if (__Pyx_GetException(&__pyx_t_27, &__pyx_t_7, &__pyx_t_2) < 0) __PYX_ERR(0, 511, __pyx_L65_except_error) __Pyx_XGOTREF(__pyx_t_27); __Pyx_XGOTREF(__pyx_t_7); __Pyx_XGOTREF(__pyx_t_2);
+512: py_list.append(raw.decode('utf-8'))
__pyx_t_6 = __Pyx_decode_cpp_string(__pyx_v_raw, 0, PY_SSIZE_T_MAX, NULL, NULL, PyUnicode_DecodeUTF8); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 512, __pyx_L65_except_error) __Pyx_GOTREF(__pyx_t_6); __pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_list, __pyx_t_6); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 512, __pyx_L65_except_error) __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0; __Pyx_XDECREF(__pyx_t_27); __pyx_t_27 = 0; __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0; __Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0; goto __pyx_L64_exception_handled; } goto __pyx_L65_except_error;
513: else:
514: # Arrays not requested, return as string
+515: py_list.append(raw.decode('utf-8'))
/*else*/ {
__pyx_t_2 = __Pyx_decode_cpp_string(__pyx_v_raw, 0, PY_SSIZE_T_MAX, NULL, NULL, PyUnicode_DecodeUTF8); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 515, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_2);
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_list, __pyx_t_2); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 515, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
}
__pyx_L60:;
516:
+517: elif first == '{':
break;
default:
518: # This is an object - always return as JSONB (bytes)
+519: py_obj = <object>PyBytes_FromStringAndSize(<const char*>raw.data(), <Py_ssize_t>raw.size())
__pyx_t_2 = PyBytes_FromStringAndSize(((char const *)__pyx_v_raw.data()), ((Py_ssize_t)__pyx_v_raw.size())); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 519, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_2); __pyx_t_7 = __pyx_t_2; __Pyx_INCREF(__pyx_t_7); __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0; __Pyx_XDECREF_SET(__pyx_v_py_obj, __pyx_t_7); __pyx_t_7 = 0;
+520: if py_obj is not None:
__pyx_t_1 = (__pyx_v_py_obj != Py_None);
if (__pyx_t_1) {
/* … */
goto __pyx_L73;
}
+521: py_list.append(py_obj)
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_list, __pyx_v_py_obj); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 521, __pyx_L1_error)
522: else:
+523: py_list.append(b'{}')
/*else*/ {
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_list, __pyx_mstate_global->__pyx_kp_b__12); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 523, __pyx_L1_error)
}
__pyx_L73:;
524: else:
525: # Unexpected - fallback to bytes
+526: py_obj = <object>PyBytes_FromStringAndSize(<const char*>raw.data(), <Py_ssize_t>raw.size())
__pyx_t_7 = PyBytes_FromStringAndSize(((char const *)__pyx_v_raw.data()), ((Py_ssize_t)__pyx_v_raw.size())); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 526, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_7); __pyx_t_2 = __pyx_t_7; __Pyx_INCREF(__pyx_t_2); __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0; __Pyx_XDECREF_SET(__pyx_v_py_obj, __pyx_t_2); __pyx_t_2 = 0;
+527: if py_obj is not None:
__pyx_t_1 = (__pyx_v_py_obj != Py_None);
if (__pyx_t_1) {
/* … */
goto __pyx_L74;
}
+528: py_list.append(py_obj)
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_list, __pyx_v_py_obj); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 528, __pyx_L1_error)
529: else:
+530: py_list.append(b'')
/*else*/ {
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_list, __pyx_mstate_global->__pyx_kp_b_); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 530, __pyx_L1_error)
}
__pyx_L74:;
break;
}
__pyx_L56_continue:;
}
+531: py_columns.append(py_list)
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_columns, __pyx_v_py_list); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 531, __pyx_L1_error)
+532: elif col_type == 'boolean':
__pyx_t_1 = (__Pyx_PyUnicode_Equals(__pyx_v_col_type, __pyx_mstate_global->__pyx_n_u_boolean, Py_EQ)); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 532, __pyx_L1_error) if (__pyx_t_1) { /* … */ goto __pyx_L16; }
+533: py_list = []
__pyx_t_2 = PyList_New(0); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 533, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_2); __Pyx_XDECREF_SET(__pyx_v_py_list, ((PyObject*)__pyx_t_2)); __pyx_t_2 = 0;
+534: for j in range(col.boolean_values.size()):
__pyx_t_28 = __pyx_v_col->boolean_values.size();
__pyx_t_29 = __pyx_t_28;
for (__pyx_t_12 = 0; __pyx_t_12 < __pyx_t_29; __pyx_t_12+=1) {
__pyx_v_j = __pyx_t_12;
+535: if col.null_mask[j]:
__pyx_t_1 = ((__pyx_v_col->null_mask[__pyx_v_j]) != 0);
if (__pyx_t_1) {
/* … */
goto __pyx_L77;
}
+536: py_list.append(None)
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_list, Py_None); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 536, __pyx_L1_error)
537: else:
+538: py_list.append(bool(col.boolean_values[j]))
/*else*/ {
__pyx_t_2 = __Pyx_PyLong_From_uint8_t((__pyx_v_col->boolean_values[__pyx_v_j])); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 538, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_2);
__pyx_t_1 = __Pyx_PyObject_IsTrue(__pyx_t_2); if (unlikely((__pyx_t_1 < 0))) __PYX_ERR(0, 538, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
__pyx_t_2 = __Pyx_PyBool_FromLong((!(!__pyx_t_1))); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 538, __pyx_L1_error)
__Pyx_GOTREF(__pyx_t_2);
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_list, __pyx_t_2); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 538, __pyx_L1_error)
__Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
}
__pyx_L77:;
}
+539: py_columns.append(py_list)
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_columns, __pyx_v_py_list); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 539, __pyx_L1_error)
540: else:
+541: py_columns.append(None)
/*else*/ {
__pyx_t_14 = __Pyx_PyList_Append(__pyx_v_py_columns, Py_None); if (unlikely(__pyx_t_14 == ((int)-1))) __PYX_ERR(0, 541, __pyx_L1_error)
}
__pyx_L16:;
__pyx_L13_continue:;
}
+542: return {
__Pyx_XDECREF(__pyx_r);
+543: 'success': True,
__pyx_t_2 = __Pyx_PyDict_NewPresized(4); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 543, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_2); if (PyDict_SetItem(__pyx_t_2, __pyx_mstate_global->__pyx_n_u_success, Py_True) < (0)) __PYX_ERR(0, 543, __pyx_L1_error)
+544: 'column_names': py_column_names,
if (PyDict_SetItem(__pyx_t_2, __pyx_mstate_global->__pyx_n_u_column_names, __pyx_v_py_column_names) < (0)) __PYX_ERR(0, 543, __pyx_L1_error)
+545: 'num_rows': table.num_rows,
__pyx_t_7 = __Pyx_PyLong_FromSize_t(__pyx_v_table.num_rows); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 545, __pyx_L1_error) __Pyx_GOTREF(__pyx_t_7); if (PyDict_SetItem(__pyx_t_2, __pyx_mstate_global->__pyx_n_u_num_rows, __pyx_t_7) < (0)) __PYX_ERR(0, 543, __pyx_L1_error) __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
+546: 'columns': py_columns
if (PyDict_SetItem(__pyx_t_2, __pyx_mstate_global->__pyx_n_u_columns, __pyx_v_py_columns) < (0)) __PYX_ERR(0, 543, __pyx_L1_error) __pyx_r = __pyx_t_2; __pyx_t_2 = 0; goto __pyx_L0;
547: }