X-Git-Url: https://pintos-os.org/cgi-bin/gitweb.cgi?a=blobdiff_plain;ds=sidebyside;f=src%2Flanguage%2Flexer%2Flexer.c;h=b6cab8d6810e4479c9faff807388857809ad0e53;hb=ae6d431159a39294af19de581292d5a21a47d557;hp=e72a3e47bc9637d93ad3c1f764b7099686df6d3d;hpb=a3368f3a6ecbc0ec5aa1212efe534288b277f13c;p=pspp diff --git a/src/language/lexer/lexer.c b/src/language/lexer/lexer.c index e72a3e47bc..b6cab8d681 100644 --- a/src/language/lexer/lexer.c +++ b/src/language/lexer/lexer.c @@ -1,5 +1,5 @@ /* PSPP - a program for statistical analysis. - Copyright (C) 1997-9, 2000, 2006, 2009, 2010, 2011 Free Software Foundation, Inc. + Copyright (C) 1997-9, 2000, 2006, 2009, 2010, 2011, 2013 Free Software Foundation, Inc. This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by @@ -30,7 +30,6 @@ #include #include -#include "data/file-name.h" #include "language/command.h" #include "language/lexer/scan.h" #include "language/lexer/segment.h" @@ -129,8 +128,9 @@ lex_reader_init (struct lex_reader *reader, { reader->class = class; reader->syntax = LEX_SYNTAX_AUTO; - reader->error = LEX_ERROR_INTERACTIVE; + reader->error = LEX_ERROR_CONTINUE; reader->file_name = NULL; + reader->encoding = NULL; reader->line_number = 0; } @@ -830,99 +830,58 @@ lex_next_tokss (const struct lexer *lexer, int n) return lex_next (lexer, n)->string; } -/* If LEXER is positioned at the (pseudo)identifier S, skips it and returns - true. Otherwise, returns false. - - S may consist of an arbitrary number of identifiers, integers, and - punctuation e.g. "KRUSKAL-WALLIS", "2SLS", or "END INPUT PROGRAM". - Identifiers may be abbreviated to their first three letters. Currently only - hyphens, slashes, and equals signs are supported as punctuation (but it - would be easy to add more). - - S must be an ASCII string. */ -bool -lex_match_phrase (struct lexer *lexer, const char *s) +static bool +lex_tokens_match (const struct token *actual, const struct token *expected) { - int tok_idx; + if (actual->type != expected->type) + return false; - for (tok_idx = 0; ; tok_idx++) + switch (actual->type) { - enum token_type token; - unsigned char c; + case T_POS_NUM: + case T_NEG_NUM: + return actual->number == expected->number; - while (c_isspace (*s)) - s++; + case T_ID: + return lex_id_match (expected->string, actual->string); - c = *s; - if (c == '\0') - { - int i; - - for (i = 0; i < tok_idx; i++) - lex_get (lexer); - return true; - } + case T_STRING: + return (actual->string.length == expected->string.length + && !memcmp (actual->string.string, expected->string.string, + actual->string.length)); - token = lex_next_token (lexer, tok_idx); - switch (c) - { - case '-': - if (token != T_DASH) - return false; - s++; - break; - - case '/': - if (token != T_SLASH) - return false; - s++; - break; - - case '=': - if (token != T_EQUALS) - return false; - s++; - break; - - case '0': case '1': case '2': case '3': case '4': - case '5': case '6': case '7': case '8': case '9': - { - unsigned int value; - - if (token != T_POS_NUM) - return false; - - value = 0; - do - { - value = value * 10 + (*s++ - '0'); - } - while (c_isdigit (*s)); - - if (lex_next_tokval (lexer, tok_idx) != value) - return false; - } - break; + default: + return true; + } +} - default: - if (lex_is_id1 (c)) - { - int len; +/* If LEXER is positioned at the sequence of tokens that may be parsed from S, + skips it and returns true. Otherwise, returns false. - if (token != T_ID) - return false; + S may consist of an arbitrary sequence of tokens, e.g. "KRUSKAL-WALLIS", + "2SLS", or "END INPUT PROGRAM". Identifiers may be abbreviated to their + first three letters. */ +bool +lex_match_phrase (struct lexer *lexer, const char *s) +{ + struct string_lexer slex; + struct token token; + int i; - len = lex_id_get_length (ss_cstr (s)); - if (!lex_id_match (ss_buffer (s, len), - lex_next_tokss (lexer, tok_idx))) - return false; + i = 0; + string_lexer_init (&slex, s, SEG_MODE_INTERACTIVE); + while (string_lexer_next (&slex, &token)) + if (token.type != SCAN_SKIP) + { + bool match = lex_tokens_match (lex_next (lexer, i++), &token); + token_destroy (&token); + if (!match) + return false; + } - s += len; - } - else - NOT_REACHED (); - } - } + while (i-- > 0) + lex_get (lexer); + return true; } static int @@ -1079,6 +1038,14 @@ lex_get_file_name (const struct lexer *lexer) return src == NULL ? NULL : src->reader->file_name; } +const char * +lex_get_encoding (const struct lexer *lexer) +{ + struct lex_source *src = lex_source__ (lexer); + return src == NULL ? NULL : src->reader->encoding; +} + + /* Returns the syntax mode for the syntax file from which the current drawn is drawn. Returns LEX_SYNTAX_AUTO for a T_STOP token or if the command's source does not have line numbers. @@ -1094,7 +1061,7 @@ lex_get_syntax_mode (const struct lexer *lexer) } /* Returns the error mode for the syntax file from which the current drawn is - drawn. Returns LEX_ERROR_INTERACTIVE for a T_STOP token or if the command's + drawn. Returns LEX_ERROR_TERMINAL for a T_STOP token or if the command's source does not have line numbers. There is no version of this function that takes an N argument because @@ -1104,13 +1071,12 @@ enum lex_error_mode lex_get_error_mode (const struct lexer *lexer) { struct lex_source *src = lex_source__ (lexer); - return src == NULL ? LEX_ERROR_INTERACTIVE : src->reader->error; + return src == NULL ? LEX_ERROR_TERMINAL : src->reader->error; } /* If the source that LEXER is currently reading has error mode - LEX_ERROR_INTERACTIVE, discards all buffered input and tokens, so that the - next token to be read comes directly from whatever is next read from the - stream. + LEX_ERROR_TERMINAL, discards all buffered input and tokens, so that the next + token to be read comes directly from whatever is next read from the stream. It makes sense to call this function after encountering an error in a command entered on the console, because usually the user would prefer not to @@ -1119,7 +1085,7 @@ void lex_interactive_reset (struct lexer *lexer) { struct lex_source *src = lex_source__ (lexer); - if (src != NULL && src->reader->error == LEX_ERROR_INTERACTIVE) + if (src != NULL && src->reader->error == LEX_ERROR_TERMINAL) { src->head = src->tail = 0; src->journal_pos = src->seg_pos = src->line_pos = 0; @@ -1141,7 +1107,7 @@ lex_discard_rest_of_command (struct lexer *lexer) } /* Discards all lookahead tokens in LEXER, then discards all input sources - until it encounters one with error mode LEX_ERROR_INTERACTIVE or until it + until it encounters one with error mode LEX_ERROR_TERMINAL or until it runs out of input sources. */ void lex_discard_noninteractive (struct lexer *lexer) @@ -1153,7 +1119,7 @@ lex_discard_noninteractive (struct lexer *lexer) while (!deque_is_empty (&src->deque)) lex_source_pop__ (src); - for (; src != NULL && src->reader->error != LEX_ERROR_INTERACTIVE; + for (; src != NULL && src->reader->error != LEX_ERROR_TERMINAL; src = lex_source__ (lexer)) lex_source_destroy (src); } @@ -1209,14 +1175,18 @@ lex_source_read__ (struct lex_source *src) do { size_t head_ofs; + size_t space; size_t n; lex_source_expand__ (src); head_ofs = src->head - src->tail; + space = src->allocated - head_ofs; n = src->reader->class->read (src->reader, &src->buffer[head_ofs], - src->allocated - head_ofs, + space, segmenter_get_prompt (&src->segmenter)); + assert (n <= space); + if (n == 0) { /* End of input. @@ -1565,9 +1535,11 @@ static void lex_source_destroy (struct lex_source *src) { char *file_name = src->reader->file_name; + char *encoding = src->reader->encoding; if (src->reader->class->destroy != NULL) src->reader->class->destroy (src->reader); free (file_name); + free (encoding); free (src->buffer); while (!deque_is_empty (&src->deque)) lex_source_pop__ (src); @@ -1580,7 +1552,6 @@ struct lex_file_reader { struct lex_reader reader; struct u8_istream *istream; - char *file_name; }; static struct lex_reader_class lex_file_reader_class; @@ -1614,9 +1585,9 @@ lex_reader_for_file (const char *file_name, const char *encoding, r->reader.syntax = syntax; r->reader.error = error; r->reader.file_name = xstrdup (file_name); + r->reader.encoding = encoding ? xstrdup (encoding) : NULL; r->reader.line_number = 1; r->istream = istream; - r->file_name = xstrdup (file_name); return &r->reader; } @@ -1635,7 +1606,7 @@ lex_file_read (struct lex_reader *r_, char *buf, size_t n, ssize_t n_read = u8_istream_read (r->istream, buf, n); if (n_read < 0) { - msg (ME, _("Error reading `%s': %s."), r->file_name, strerror (errno)); + msg (ME, _("Error reading `%s': %s."), r_->file_name, strerror (errno)); return 0; } return n_read; @@ -1649,12 +1620,11 @@ lex_file_close (struct lex_reader *r_) if (u8_istream_fileno (r->istream) != STDIN_FILENO) { if (u8_istream_close (r->istream) != 0) - msg (ME, _("Error closing `%s': %s."), r->file_name, strerror (errno)); + msg (ME, _("Error closing `%s': %s."), r_->file_name, strerror (errno)); } else u8_istream_free (r->istream); - free (r->file_name); free (r); } @@ -1674,16 +1644,17 @@ struct lex_string_reader static struct lex_reader_class lex_string_reader_class; /* Creates and returns a new lex_reader for the contents of S, which must be - encoded in UTF-8. The new reader takes ownership of S and will free it + encoded in the given ENCODING. The new reader takes ownership of S and will free it with ss_dealloc() when it is closed. */ struct lex_reader * -lex_reader_for_substring_nocopy (struct substring s) +lex_reader_for_substring_nocopy (struct substring s, const char *encoding) { struct lex_string_reader *r; r = xmalloc (sizeof *r); lex_reader_init (&r->reader, &lex_string_reader_class); - r->reader.syntax = LEX_SYNTAX_INTERACTIVE; + r->reader.syntax = LEX_SYNTAX_AUTO; + r->reader.encoding = encoding ? xstrdup (encoding) : NULL; r->s = s; r->offset = 0; @@ -1691,25 +1662,25 @@ lex_reader_for_substring_nocopy (struct substring s) } /* Creates and returns a new lex_reader for a copy of null-terminated string S, - which must be encoded in UTF-8. The caller retains ownership of S. */ + which must be encoded in ENCODING. The caller retains ownership of S. */ struct lex_reader * -lex_reader_for_string (const char *s) +lex_reader_for_string (const char *s, const char *encoding) { struct substring ss; ss_alloc_substring (&ss, ss_cstr (s)); - return lex_reader_for_substring_nocopy (ss); + return lex_reader_for_substring_nocopy (ss, encoding); } /* Formats FORMAT as a printf()-like format string and creates and returns a new lex_reader for the formatted result. */ struct lex_reader * -lex_reader_for_format (const char *format, ...) +lex_reader_for_format (const char *format, const char *encoding, ...) { struct lex_reader *r; va_list args; - va_start (args, format); - r = lex_reader_for_substring_nocopy (ss_cstr (xvasprintf (format, args))); + va_start (args, encoding); + r = lex_reader_for_substring_nocopy (ss_cstr (xvasprintf (format, args)), encoding); va_end (args); return r;