cvs: php-src /ext/standard file.c

pollita Tue May 2 18:23:33 2006 UTC Modified files: /php-src/ext/standard file.c Log: Unicode Updates: file() now unicode ready get_meta_tags() updated for FS encoding, but content is still non-unicode http://cvs.php.net/viewcvs.cgi/php-src/ext/standard/file.c?r1=1.439&r2=1.440&diff_format=u Index: php-src/ext/standard/file.c diff -u php-src/ext/standard/file.c:1.439 php-src/ext/standard/file.c:1.440 --- php-src/ext/standard/file.c:1.439 Fri Apr 28 19:03:57 2006 +++ php-src/ext/standard/file.c Tue May 2 18:23:32 2006 @@ -21,7 +21,7 @@ +----------------------------------------------------------------------+ */ -/* $Id: file.c,v 1.439 2006/04/28 19:03:57 fmk Exp $ */ +/* $Id: file.c,v 1.440 2006/05/02 18:23:32 pollita Exp $ */ /* Synced with php 3.0 revision 1.218 1999-06-16 [ssb] */ @@ -360,6 +360,7 @@ { char *filename; int filename_len; + zend_uchar filename_type; zend_bool use_include_path = 0; int in_tag = 0, done = 0; int looking_for_val = 0, have_name = 0, have_content = 0; @@ -372,15 +373,22 @@ memset(&md, 0, sizeof(md)); /* Parse arguments */ - if (zend_parse_parameters(ZEND_NUM_ARGS() TSRMLS_CC, "s|b", - &filename, &filename_len, &use_include_path) == FAILURE) { + if (zend_parse_parameters(ZEND_NUM_ARGS() TSRMLS_CC, "t|b", + &filename, &filename_len, &filename_type, &use_include_path) == FAILURE) { return; } + if (filename_type == IS_UNICODE) { + if (php_stream_path_encode(NULL, &filename, &filename_len, (UChar*)filename, filename_len, REPORT_ERRORS, FG(default_context)) == FAILURE) { + RETURN_FALSE; + } + } md.stream = php_stream_open_wrapper(filename, "rb", (use_include_path ? USE_PATH : 0) | REPORT_ERRORS, NULL); - + if (filename_type == IS_UNICODE) { + efree(filename); + } if (!md.stream) { RETURN_FALSE; } @@ -506,7 +514,6 @@ zend_uchar filename_type; void *contents; long flags = 0; - zend_bool use_include_path = 0; php_stream *stream; int len; long offset = -1; @@ -732,7 +739,7 @@ } /* }}} */ -/* {{{ proto array file(string filename [, int flags[, resource context]]) +/* {{{ proto array file(string filename [, int flags[, resource context]]) U Read entire file into an array */ #define PHP_FILE_BUF_SIZE 80 @@ -742,23 +749,24 @@ { char *filename; int filename_len; - char *target_buf=NULL, *p, *s, *e; + zend_uchar filename_type; + char *target_buf=NULL; register int i = 0; int target_len; - char eol_marker = '\n'; long flags = 0; zend_bool use_include_path; zend_bool include_new_line; zend_bool skip_blank_lines; + zend_bool text_mode; php_stream *stream; zval *zcontext = NULL; php_stream_context *context = NULL; /* Parse arguments */ - if (zend_parse_parameters(ZEND_NUM_ARGS() TSRMLS_CC, "s|lr!", &filename, &filename_len, &flags, &zcontext) == FAILURE) { + if (zend_parse_parameters(ZEND_NUM_ARGS() TSRMLS_CC, "t|lr!", &filename, &filename_len, &filename_type, &flags, &zcontext) == FAILURE) { return; } - if (flags < 0 || flags > (PHP_FILE_USE_INCLUDE_PATH | PHP_FILE_IGNORE_NEW_LINES | PHP_FILE_SKIP_EMPTY_LINES | PHP_FILE_NO_DEFAULT_CONTEXT)) { + if (flags < 0 || flags > (PHP_FILE_USE_INCLUDE_PATH | PHP_FILE_IGNORE_NEW_LINES | PHP_FILE_SKIP_EMPTY_LINES | PHP_FILE_NO_DEFAULT_CONTEXT | PHP_FILE_TEXT)) { php_error_docref(NULL TSRMLS_CC, E_WARNING, "'%ld' flag is not supported", flags); RETURN_FALSE; } @@ -766,10 +774,19 @@ use_include_path = flags & PHP_FILE_USE_INCLUDE_PATH; include_new_line = !(flags & PHP_FILE_IGNORE_NEW_LINES); skip_blank_lines = flags & PHP_FILE_SKIP_EMPTY_LINES; + text_mode = flags & PHP_FILE_TEXT; context = php_stream_context_from_zval(zcontext, flags & PHP_FILE_NO_DEFAULT_CONTEXT); - stream = php_stream_open_wrapper_ex(filename, "rb", (use_include_path ? USE_PATH : 0) | REPORT_ERRORS, NULL, context); + if (filename_type == IS_UNICODE) { + if (php_stream_path_encode(NULL, &filename, &filename_len, (UChar*)filename, filename_len, REPORT_ERRORS, context) == FAILURE) { + RETURN_FALSE; + } + } + stream = php_stream_open_wrapper_ex(filename, text_mode ? "rt" : "rb", (use_include_path ? USE_PATH : 0) | REPORT_ERRORS, NULL, context); + if (filename_type == IS_UNICODE) { + efree(filename); + } if (!stream) { RETURN_FALSE; } @@ -777,9 +794,53 @@ /* Initialize return array */ array_init(return_value); - if ((target_len = php_stream_copy_to_mem(stream, &target_buf, PHP_STREAM_COPY_ALL, 0))) { - s = target_buf; - e = target_buf + target_len; + target_len = php_stream_copy_to_mem_ex(stream, stream->readbuf_type, (void**)&target_buf, PHP_STREAM_COPY_ALL, -1, 0); + + if (!target_len) { + /* Empty file, do nothing and return an empty array */ + } else if (stream->readbuf_type == IS_UNICODE) { + UChar *s = (UChar*)target_buf, *p; + UChar *e = s + target_len, eol_marker = '\n'; + + if (!(p = php_stream_locate_eol(stream, ZSTR(target_buf), target_len TSRMLS_CC))) { + p = e; + goto uparse_eol; + } + + if (stream->flags & PHP_STREAM_FLAG_EOL_MAC) { + eol_marker = '\r'; + } + + /* for performance reasons the code is quadruplicated, so that the if (include_new_line/unicode + * will not need to be done for every single line in the file. + */ + if (include_new_line) { + do { + p++; +uparse_eol: + add_index_unicodel(return_value, i++, eustrndup(s, p-s), p-s, 0); + s = p; + } while ((p = u_memchr(p, eol_marker, (e-p)))); + } else { + do { + if (skip_blank_lines && !(p-s)) { + s = ++p; + continue; + } + add_index_unicodel(return_value, i++, eustrndup(s, p-s), p-s, 0); + s = ++p; + } while ((p = u_memchr(p, eol_marker, (e-p)))); + } + + /* handle any left overs of files without new lines */ + if (s != e) { + p = e; + goto uparse_eol; + } + + } else { /* !IS_UNICODE */ + char *s = target_buf, *p; + char *e = target_buf + target_len, eol_marker = '\n'; if (!(p = php_stream_locate_eol(stream, ZSTR(target_buf), target_len TSRMLS_CC))) { p = e; @@ -790,9 +851,6 @@ eol_marker = '\r'; } - /* for performance reasons the code is duplicated, so that the if (include_new_line) - * will not need to be done for every single line in the file. - */ if (include_new_line) { do { p++;