From 0f0599ddc1ad02f392fe0be58479f34c57a80c8d Mon Sep 17 00:00:00 2001 From: Andrew McNamara Date: Wed, 12 Jan 2005 09:45:18 +0000 Subject: When quoting=QUOTE_NONNUMERIC, the reader now casts unquoted fields to floats. --- Lib/test/test_csv.py | 6 ++++ Misc/NEWS | 7 +++-- Modules/_csv.c | 89 ++++++++++++++++++++++++++++++++++------------------ 3 files changed, 70 insertions(+), 32 deletions(-) diff --git a/Lib/test/test_csv.py b/Lib/test/test_csv.py index ed10ed7..d458332 100644 --- a/Lib/test/test_csv.py +++ b/Lib/test/test_csv.py @@ -241,6 +241,12 @@ class Test_Csv(unittest.TestCase): quotechar=None, escapechar='\\') self._read_test(['1,",3,",5'], [['1', '"', '3', '"', '5']], quoting=csv.QUOTE_NONE, escapechar='\\') + # will this fail where locale uses comma for decimals? + self._read_test([',3,"5",7.3'], [['', 3, '5', 7.3]], + quoting=csv.QUOTE_NONNUMERIC) + self.assertRaises(ValueError, self._read_test, + ['abc,3'], [[]], + quoting=csv.QUOTE_NONNUMERIC) def test_read_bigfield(self): # This exercises the buffer realloc functionality and field size diff --git a/Misc/NEWS b/Misc/NEWS index 02f54bd..828063a 100644 --- a/Misc/NEWS +++ b/Misc/NEWS @@ -45,8 +45,11 @@ Library + quotechar=None and quoting=QUOTE_NONE now work the way PEP 305 dictates. + the parser now removes the escapechar prefix from escaped characters. - + QUOTE_NONNUMERIC now tests for numeric objects, rather than attempting - to cast to float. + + when quoting=QUOTE_NONNUMERIC, the writer now tests for numeric + objects, rather than attempting to cast to float, and using the + success of that as the determinator. + + when quoting=QUOTE_NONNUMERIC, the reader now casts unquoted fields + to floats. + writer doublequote handling improved. + Dialect classes passed to the module are no longer instantiated by the module before being parsed (the former validation scheme required diff --git a/Modules/_csv.c b/Modules/_csv.c index c592933..6941714 100644 --- a/Modules/_csv.c +++ b/Modules/_csv.c @@ -97,6 +97,7 @@ typedef struct { int field_size; /* size of allocated buffer */ int field_len; /* length of current field */ int had_parse_error; /* did we have a parse error? */ + int numeric_field; /* treat field as numeric */ } ReaderObj; staticforward PyTypeObject Reader_Type; @@ -495,17 +496,30 @@ _call_dialect(PyObject *dialect_inst, PyObject *kwargs) return dialect; } -static void +static int parse_save_field(ReaderObj *self) { PyObject *field; field = PyString_FromStringAndSize(self->field, self->field_len); - if (field != NULL) { - PyList_Append(self->fields, field); - Py_XDECREF(field); - } + if (field == NULL) + return -1; self->field_len = 0; + if (self->numeric_field) { + PyObject *tmp; + + self->numeric_field = 0; + tmp = PyNumber_Float(field); + if (tmp == NULL) { + Py_DECREF(field); + return -1; + } + Py_DECREF(field); + field = tmp; + } + PyList_Append(self->fields, field); + Py_DECREF(field); + return 0; } static int @@ -529,6 +543,22 @@ parse_grow_buff(ReaderObj *self) } static int +parse_reset(ReaderObj *self) +{ + if (self->fields) { + Py_DECREF(self->fields); + } + self->fields = PyList_New(0); + if (self->fields == NULL) + return -1; + self->field_len = 0; + self->state = START_RECORD; + self->had_parse_error = 0; + self->numeric_field = 0; + return 0; +} + +static int parse_add_char(ReaderObj *self, char c) { if (self->field_len >= field_limit) { @@ -560,7 +590,8 @@ parse_process_char(ReaderObj *self, char c) /* expecting field */ if (c == '\n') { /* save empty field - return [fields] */ - parse_save_field(self); + if (parse_save_field(self) < 0) + return -1; self->state = START_RECORD; } else if (c == dialect->quotechar && @@ -577,10 +608,13 @@ parse_process_char(ReaderObj *self, char c) ; else if (c == dialect->delimiter) { /* save empty field */ - parse_save_field(self); + if (parse_save_field(self) < 0) + return -1; } else { /* begin new unquoted field */ + if (dialect->quoting == QUOTE_NONNUMERIC) + self->numeric_field = 1; if (parse_add_char(self, c) < 0) return -1; self->state = IN_FIELD; @@ -597,7 +631,8 @@ parse_process_char(ReaderObj *self, char c) /* in unquoted field */ if (c == '\n') { /* end of line - return [fields] */ - parse_save_field(self); + if (parse_save_field(self) < 0) + return -1; self->state = START_RECORD; } else if (c == dialect->escapechar) { @@ -606,7 +641,8 @@ parse_process_char(ReaderObj *self, char c) } else if (c == dialect->delimiter) { /* save field - wait for new field */ - parse_save_field(self); + if (parse_save_field(self) < 0) + return -1; self->state = START_FIELD; } else { @@ -662,12 +698,14 @@ parse_process_char(ReaderObj *self, char c) } else if (c == dialect->delimiter) { /* save field - wait for new field */ - parse_save_field(self); + if (parse_save_field(self) < 0) + return -1; self->state = START_FIELD; } else if (c == '\n') { /* end of line - return [fields] */ - parse_save_field(self); + if (parse_save_field(self) < 0) + return -1; self->state = START_RECORD; } else if (!dialect->strict) { @@ -716,15 +754,11 @@ Reader_iternext(ReaderObj *self) return NULL; } - if (self->had_parse_error) { - if (self->fields) { - Py_XDECREF(self->fields); - } - self->fields = PyList_New(0); - self->field_len = 0; - self->state = START_RECORD; - self->had_parse_error = 0; - } + if (self->had_parse_error) + if (parse_reset(self) < 0) { + Py_DECREF(lineobj); + return NULL; + } line = PyString_AsString(lineobj); if (line == NULL) { @@ -886,15 +920,15 @@ csv_reader(PyObject *module, PyObject *args, PyObject *keyword_args) return NULL; self->dialect = NULL; - self->input_iter = self->fields = NULL; - self->fields = NULL; self->input_iter = NULL; - self->had_parse_error = 0; self->field = NULL; self->field_size = 0; - self->field_len = 0; - self->state = START_RECORD; + + if (parse_reset(self) < 0) { + Py_DECREF(self); + return NULL; + } if (!PyArg_UnpackTuple(args, "", 1, 2, &iterator, &dialect)) { Py_DECREF(self); @@ -912,11 +946,6 @@ csv_reader(PyObject *module, PyObject *args, PyObject *keyword_args) Py_DECREF(self); return NULL; } - self->fields = PyList_New(0); - if (self->fields == NULL) { - Py_DECREF(self); - return NULL; - } PyObject_GC_Track(self); return (PyObject *)self; -- cgit v0.12