smarter_csv 1.14.3 → 1.14.4

data/ext/smarter_csv/smarter_csv.c.works9

@@ -1,196 +0,0 @@
-#include "ruby.h"
-#include "ruby/encoding.h"
-#include <stdio.h>
-#include <stdbool.h>
-#include <string.h>
-
-#ifndef bool
-  #define bool int
-  #define false ((bool)0)
-  #define true  ((bool)1)
-#endif
-
-VALUE SmarterCSV = Qnil;
-VALUE eMalformedCSVError = Qnil;
-VALUE Parser = Qnil;
-VALUE Qempty_string = Qnil; // shared frozen empty string
-
-typedef struct {
-  const char *start;
-  long length;
-  bool needs_unescape;
-} ParsedField;
-
-static bool detect_and_strip_quotes(const char **start, long *len, char quote_char, bool *found_double) {
-  if (*len < 2 || (*start)[0] != quote_char || (*start)[*len - 1] != quote_char) return false;
-
-  *start = *start + 1;
-  *len = *len - 2;
-
-  for (long i = 0; i < *len - 1; i++) {
-    if ((*start)[i] == quote_char && (*start)[i + 1] == quote_char) {
-      *found_double = true;
-      return true;
-    }
-  }
-  *found_double = false;
-  return true;
-}
-
-static VALUE unescape_quotes(const char *start, long len, char quote_char, rb_encoding *enc) {
-  char *buf = ALLOC_N(char, len);
-  long j = 0;
-
-  for (long i = 0; i < len; i++) {
-    if (start[i] == quote_char && i + 1 < len && start[i + 1] == quote_char) {
-      buf[j++] = quote_char;
-      i++;
-    } else {
-      buf[j++] = start[i];
-    }
-  }
-
-  VALUE str = rb_enc_str_new(buf, j, enc);
-  xfree(buf);
-  return str;
-}
-
-static VALUE rb_parse_csv_line(VALUE self, VALUE line, VALUE col_sep, VALUE quote_char, VALUE max_size, VALUE has_quotes_val) {
-  if (RB_TYPE_P(line, T_NIL)) return rb_ary_new();
-  if (!RB_TYPE_P(line, T_STRING)) rb_raise(rb_eTypeError, "ERROR in SmarterCSV.parse_line: line has to be a string or nil");
-
-  rb_encoding *encoding = rb_enc_get(line);
-  char *startP = RSTRING_PTR(line);
-  long line_len = RSTRING_LEN(line);
-  char *endP = startP + line_len;
-  char *p = startP;
-
-  char *col_sepP = RSTRING_PTR(col_sep);
-  long col_sep_len = RSTRING_LEN(col_sep);
-
-  char *quoteP = RSTRING_PTR(quote_char);
-  char quote_char_val = quoteP[0];
-  size_t quote_len = strlen(quoteP);
-
-  VALUE elements = rb_ary_new();
-
-  long backslash_count = 0;
-  bool in_quotes = false;
-  long element_count = 0;
-
-  int max_fields = -1;
-  if (max_size != Qnil) {
-    max_fields = NUM2INT(max_size);
-    if (max_fields < 0) return rb_ary_new();
-  }
-
-  bool has_quotes = RTEST(has_quotes_val);
-
-  if (__builtin_expect(!has_quotes && col_sep_len == 1, 1)) {
-    char sep = *col_sepP;
-    char *sep_pos = NULL;
-
-    while ((sep_pos = memchr(p, sep, endP - p))) {
-      if ((max_fields >= 0) && (element_count >= max_fields)) break;
-
-      bool only_spaces = true;
-      for (char *s = startP; s < sep_pos; s++) {
-        if (*s != ' ') { only_spaces = false; break; }
-      }
-
-      VALUE field = only_spaces ? Qempty_string : rb_enc_str_new(startP, sep_pos - startP, encoding);
-      rb_ary_push(elements, field);
-      element_count++;
-
-      p = sep_pos + 1;
-      startP = p;
-    }
-
-    if ((max_fields < 0) || (element_count < max_fields)) {
-      bool only_spaces = true;
-      for (char *s = startP; s < endP; s++) {
-        if (*s != ' ') { only_spaces = false; break; }
-      }
-      VALUE field = only_spaces ? Qempty_string : rb_enc_str_new(startP, endP - startP, encoding);
-      rb_ary_push(elements, field);
-    }
-
-    return elements;
-  }
-
-  ParsedField *fields = ALLOC_N(ParsedField, 128);
-  long field_count = 0;
-  char *field_start = startP;
-
-  while (p < endP) {
-    bool col_sep_found = (p + col_sep_len <= endP) && (memcmp(p, col_sepP, col_sep_len) == 0);
-
-    if (col_sep_found && !in_quotes) {
-      if ((max_fields >= 0) && (field_count >= max_fields)) break;
-
-      long field_len = p - field_start;
-      const char *actual_start = field_start;
-      long actual_len = field_len;
-      bool needs_unescape = false;
-      bool quoted = detect_and_strip_quotes(&actual_start, &actual_len, quote_char_val, &needs_unescape);
-      if (!quoted) {
-        for (long i = 0; i < actual_len - 1; i++) {
-          if (actual_start[i] == quote_char_val && actual_start[i + 1] == quote_char_val) {
-            needs_unescape = true;
-            break;
-          }
-        }
-      }
-
-      fields[field_count++] = (ParsedField){ actual_start, actual_len, needs_unescape };
-      p += col_sep_len;
-      field_start = p;
-      backslash_count = 0;
-    } else {
-      if (*p == '\\') {
-        backslash_count++;
-      } else {
-        if (*p == quote_char_val && (backslash_count % 2 == 0)) {
-          in_quotes = !in_quotes;
-        }
-        backslash_count = 0;
-      }
-      p++;
-    }
-  }
-
-  if (in_quotes) {
-    xfree(fields);
-    rb_raise(eMalformedCSVError, "Unclosed quoted field detected in line: %s", StringValueCStr(line));
-  }
-
-  if ((max_fields < 0) || (field_count < max_fields)) {
-    long field_len = endP - field_start;
-    const char *actual_start = field_start;
-    long actual_len = field_len;
-    bool needs_unescape = false;
-    detect_and_strip_quotes(&actual_start, &actual_len, quote_char_val, &needs_unescape);
-
-    fields[field_count++] = (ParsedField){ actual_start, actual_len, needs_unescape };
-  }
-
-  for (long i = 0; i < field_count; i++) {
-    ParsedField f = fields[i];
-    VALUE field = f.length == 0 ? Qempty_string :
-                  (f.needs_unescape ? unescape_quotes(f.start, f.length, quote_char_val, encoding)
-                                    : rb_enc_str_new(f.start, f.length, encoding));
-    rb_ary_push(elements, field);
-  }
-
-  xfree(fields);
-  return elements;
-}
-
-void Init_smarter_csv(void) {
-  SmarterCSV = rb_const_get(rb_cObject, rb_intern("SmarterCSV"));
-  Parser = rb_const_get(SmarterCSV, rb_intern("Parser"));
-  eMalformedCSVError = rb_const_get(SmarterCSV, rb_intern("MalformedCSV"));
-  Qempty_string = rb_str_new_literal("");
-  rb_gc_register_address(&Qempty_string);
-  rb_define_module_function(Parser, "parse_csv_line_c", rb_parse_csv_line, 5);
-}