X-Git-Url: https://pintos-os.org/cgi-bin/gitweb.cgi?a=blobdiff_plain;f=src%2Flanguage%2Fdata-io%2Fprint.c;h=9768c4535af36bd8962b429f7e667cd570b059d1;hb=refs%2Fheads%2Flexer;hp=e345477baf706f50add706bbac7e5f04b0d4057d;hpb=8f04b0ced35a66cfdebefbcb53c81979add36ca3;p=pspp diff --git a/src/language/data-io/print.c b/src/language/data-io/print.c index e345477baf..9768c4535a 100644 --- a/src/language/data-io/print.c +++ b/src/language/data-io/print.c @@ -1,5 +1,5 @@ /* PSPP - a program for statistical analysis. - Copyright (C) 1997-9, 2000, 2006, 2009 Free Software Foundation, Inc. + Copyright (C) 1997-9, 2000, 2006, 2009, 2010, 2011, 2012 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 @@ -18,30 +18,30 @@ #include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include - -#include "xalloc.h" +#include "data/case.h" +#include "data/dataset.h" +#include "data/data-out.h" +#include "data/format.h" +#include "data/transformations.h" +#include "data/variable.h" +#include "language/command.h" +#include "language/data-io/data-writer.h" +#include "language/data-io/file-handle.h" +#include "language/data-io/placement-parser.h" +#include "language/lexer/format-parser.h" +#include "language/lexer/lexer.h" +#include "language/lexer/variable-parser.h" +#include "libpspp/assertion.h" +#include "libpspp/compiler.h" +#include "libpspp/i18n.h" +#include "libpspp/ll.h" +#include "libpspp/message.h" +#include "libpspp/misc.h" +#include "libpspp/pool.h" +#include "output/tab.h" +#include "output/text-item.h" + +#include "gl/xalloc.h" #include "gettext.h" #define _(msgid) gettext (msgid) @@ -50,7 +50,11 @@ enum field_type { PRT_LITERAL, /* Literal string. */ - PRT_VAR /* Variable. */ + PRT_VAR, /* Variable. */ + PRT_ADVANCE, /* Skip forward N positions. */ + PRT_GOTO, /* Go to absolute position N. */ + PRT_NEWLINE, /* Output the constructed line. */ + PRT_EJECT /* Eject the output page. */ }; /* Describes how to output one field. */ @@ -59,13 +63,11 @@ struct prt_out_spec /* All fields. */ struct ll ll; /* In struct print_trns `specs' list. */ enum field_type type; /* What type of field this is. */ - int record; /* 1-based record number. */ int first_column; /* 0-based first column. */ /* PRT_VAR only. */ const struct variable *var; /* Associated variable. */ struct fmt_spec format; /* Output spec. */ - bool add_space; /* Add trailing space? */ bool sysmis_as_spaces; /* Output SYSMIS as spaces? */ /* PRT_LITERAL only. */ @@ -82,13 +84,14 @@ ll_to_prt_out_spec (struct ll *ll) struct print_trns { struct pool *pool; /* Stores related data. */ - bool eject; /* Eject page before printing? */ - bool include_prefix; /* Prefix lines with space? */ const char *encoding; /* Encoding to use for output. */ struct dfm_writer *writer; /* Output file, NULL=listing file. */ struct ll_list specs; /* List of struct prt_out_specs. */ - size_t record_cnt; /* Number of records to write. */ - struct string line; /* Output buffer. */ + struct u8_line line; /* Output buffer. */ + + int unit; /* Unit width, in bytes. */ + char one[MAX_UNIT]; /* '1' in encoding, 'unit' bytes long. */ + char space[MAX_UNIT]; /* \n in encoding, 'unit' bytes long. */ }; enum which_formats @@ -136,39 +139,51 @@ internal_cmd_print (struct lexer *lexer, struct dataset *ds, bool print_table = 0; struct print_trns *trns; struct file_handle *fh = NULL; + char *encoding = NULL; struct pool *tmp_pool; + int record_cnt = 0; /* Fill in prt to facilitate error-handling. */ trns = pool_create_container (struct print_trns, pool); trns->eject = eject; trns->writer = NULL; - trns->record_cnt = 0; ll_init (&trns->specs); - ds_init_empty (&trns->line); - ds_register_pool (&trns->line, trns->pool); + u8_line_init (&trns->line); + u8_line_register_pool (&trns->line, trns->pool); tmp_pool = pool_create_subpool (trns->pool); /* Parse the command options. */ - while (lex_token (lexer) != '/' && lex_token (lexer) != '.') + while (lex_token (lexer) != T_SLASH && lex_token (lexer) != T_ENDCMD) { if (lex_match_id (lexer, "OUTFILE")) { - lex_match (lexer, '='); + lex_match (lexer, T_EQUALS); - fh = fh_parse (lexer, FH_REF_FILE); + fh = fh_parse (lexer, FH_REF_FILE, NULL); if (fh == NULL) goto error; } + else if (lex_match_id (lexer, "ENCODING")) + { + lex_match (lexer, T_EQUALS); + if (!lex_force_string (lexer)) + goto error; + + free (encoding); + encoding = ss_xstrdup (lex_tokss (lexer)); + + lex_get (lexer); + } else if (lex_match_id (lexer, "RECORDS")) { - lex_match (lexer, '='); - lex_match (lexer, '('); + lex_match (lexer, T_EQUALS); + lex_match (lexer, T_LPAREN); if (!lex_force_int (lexer)) goto error; - trns->record_cnt = lex_integer (lexer); + record_cnt = lex_integer (lexer); lex_get (lexer); - lex_match (lexer, ')'); + lex_match (lexer, T_RPAREN); } else if (lex_match_id (lexer, "TABLE")) print_table = true; @@ -181,10 +196,6 @@ internal_cmd_print (struct lexer *lexer, struct dataset *ds, } } - /* When PRINT or PRINT EJECT writes to an external file, we - prefix each line with a space for compatibility. */ - trns->include_prefix = which_formats == PRINT && fh != NULL; - /* Parse variables and strings. */ if (!parse_specs (lexer, tmp_pool, trns, dataset_dict (ds), which_formats)) goto error; @@ -194,13 +205,13 @@ internal_cmd_print (struct lexer *lexer, struct dataset *ds, if (fh != NULL) { - trns->writer = dfm_open_writer (fh); + trns->writer = dfm_open_writer (fh, encoding); if (trns->writer == NULL) goto error; - trns->encoding = dfm_writer_get_legacy_encoding (trns->writer); + trns->encoding = dfm_writer_get_encoding (trns->writer); } else - trns->encoding = LEGACY_NATIVE; + trns->encoding = UTF8; /* Output the variable table if requested. */ if (print_table) @@ -239,20 +250,20 @@ parse_specs (struct lexer *lexer, struct pool *tmp_pool, struct print_trns *trns int record = 0; int column = 1; - if (lex_token (lexer) == '.') + if (lex_token (lexer) == T_ENDCMD) { trns->record_cnt = 1; return true; } - while (lex_token (lexer) != '.') + while (lex_token (lexer) != T_ENDCMD) { bool ok; if (!parse_record_placement (lexer, &record, &column)) return false; - if (lex_token (lexer) == T_STRING) + if (lex_is_string (lexer)) ok = parse_string_argument (lexer, trns, record, &column); else ok = parse_variable_argument (lexer, dict, trns, tmp_pool, &record, &column, @@ -260,7 +271,7 @@ parse_specs (struct lexer *lexer, struct pool *tmp_pool, struct print_trns *trns if (!ok) return 0; - lex_match (lexer, ','); + lex_match (lexer, T_COMMA); } if (trns->record_cnt != 0 && trns->record_cnt != record) @@ -280,7 +291,7 @@ parse_string_argument (struct lexer *lexer, struct print_trns *trns, int record, spec->type = PRT_LITERAL; spec->record = record; spec->first_column = *column; - ds_init_string (&spec->string, lex_tokstr (lexer)); + ds_init_substring (&spec->string, lex_tokss (lexer)); ds_register_pool (&spec->string, trns->pool); lex_get (lexer); @@ -323,9 +334,9 @@ parse_variable_argument (struct lexer *lexer, const struct dictionary *dict, &vars, &var_cnt, PV_DUPLICATE)) return false; - if (lex_is_number (lexer) || lex_token (lexer) == '(') + if (lex_is_number (lexer) || lex_token (lexer) == T_LPAREN) { - if (!parse_var_placements (lexer, tmp_pool, var_cnt, false, + if (!parse_var_placements (lexer, tmp_pool, var_cnt, FMT_FOR_OUTPUT, &formats, &format_cnt)) return false; add_space = false; @@ -334,7 +345,7 @@ parse_variable_argument (struct lexer *lexer, const struct dictionary *dict, { size_t i; - lex_match (lexer, '*'); + lex_match (lexer, T_ASTERISK); formats = pool_nmalloc (tmp_pool, var_cnt, sizeof *formats); format_cnt = var_cnt; @@ -396,8 +407,7 @@ dump_table (struct print_trns *trns, const struct file_handle *fh) int row; spec_cnt = ll_count (&trns->specs); - t = tab_create (4, spec_cnt + 1, 0); - tab_columns (t, TAB_COL_DOWN, 1); + t = tab_create (4, spec_cnt + 1); tab_box (t, TAL_1, TAL_1, TAL_0, TAL_1, 0, 0, 3, spec_cnt); tab_hline (t, TAL_2, 0, 3, 1); tab_headers (t, 0, 0, 1, 0); @@ -405,7 +415,6 @@ dump_table (struct print_trns *trns, const struct file_handle *fh) tab_text (t, 1, 0, TAB_CENTER | TAT_TITLE, _("Record")); tab_text (t, 2, 0, TAB_CENTER | TAT_TITLE, _("Columns")); tab_text (t, 3, 0, TAB_CENTER | TAT_TITLE, _("Format")); - tab_dim (t, tab_natural_dimensions, NULL); row = 1; ll_for_each (spec, struct prt_out_spec, ll, &trns->specs) { @@ -414,7 +423,7 @@ dump_table (struct print_trns *trns, const struct file_handle *fh) switch (spec->type) { case PRT_LITERAL: - tab_text_format (t, 0, row, TAB_LEFT | TAB_FIX, "\"%.*s\"", + tab_text_format (t, 0, row, TAB_LEFT | TAB_FIX, "`%.*s'", (int) ds_length (&spec->string), ds_data (&spec->string)); width = ds_length (&spec->string); @@ -435,63 +444,100 @@ dump_table (struct print_trns *trns, const struct file_handle *fh) } if (fh != NULL) - tab_title (t, ngettext ("Writing %d record to %s.", - "Writing %d records to %s.", trns->record_cnt), + tab_title (t, ngettext ("Writing %zu record to %s.", + "Writing %zu records to %s.", trns->record_cnt), trns->record_cnt, fh_get_name (fh)); else - tab_title (t, ngettext ("Writing %d record.", - "Writing %d records.", trns->record_cnt), + tab_title (t, ngettext ("Writing %zu record.", + "Writing %zu records.", trns->record_cnt), trns->record_cnt); tab_submit (t); } /* Transformation. */ -static void flush_records (struct print_trns *, int target_record, - bool *eject, int *record); +static void +print_literal (struct print_trns *trns, const struct prt_out_spec *spec) +{ + const char *s = ds_data (&spec->string); + int n = ds_length (&spec->string); + int width = u8_width (s, n, UTF8); + + u8_line_put (&trns->line, x0, x0 + width, s, n); +} + +static void +print_var (struct print_trns *trns, const struct prt_out_spec *spec, + const union value *value) +{ + int x0 = spec->first_column; + + if (!spec->sysmis_as_spaces || input->f != SYSMIS) + { + char *s = data_out (input, var_get_encoding (spec->var), &spec->format); + size_t n = strlen (s); + + x1 = ; + u8_line_put (&trns->line, x0, x1, s, n); + free (s); + } + else + { + x1 = x0 + spec->format.w; + memset (u8_line_reserve (&trns->line, x0, x1, x1 - x0), + ' ', x1 - x0); + } + + +} /* Performs the transformation inside print_trns T on case C. */ static int print_trns_proc (void *trns_, struct ccase **c, casenumber case_num UNUSED) { struct print_trns *trns = trns_; - bool eject = trns->eject; - char encoded_space = legacy_from_native (trns->encoding, ' '); - int record = 1; struct prt_out_spec *spec; - ds_clear (&trns->line); - ds_put_char (&trns->line, ' '); + u8_line_clear (&trns->line); + ll_for_each (spec, struct prt_out_spec, ll, &trns->specs) + switch (spec->type) + { + case PRT_LITERAL: + print_literal (trns, spec); + break; + + case PRT_VAR: + print_var (trns, spec, case_data (*c, spec->var)); + break; + + case PRT_NEWLINE: + if (trns->writer == NULL) + tab_output_text (TAB_FIX, ds_cstr (&trns->line.s)); + else + dfm_put_record (trns->writer, ds_data (&trns->line.s), + ds_length (&trns->line.s)); + u8_line_clear (&trns->line); + break; + + case PRT_EJECT: + text_item_submit (text_item_create (TEXT_ITEM_EJECT_PAGE, "")); + break; + } + { - flush_records (trns, spec->record, &eject, &record); + int x0 = spec->first_column; - ds_set_length (&trns->line, spec->first_column, encoded_space); if (spec->type == PRT_VAR) { - const union value *input = case_data (*c, spec->var); - char *output = ds_put_uninit (&trns->line, spec->format.w); - if (!spec->sysmis_as_spaces || input->f != SYSMIS) - data_out_legacy (input, trns->encoding, &spec->format, output); - else - memset (output, encoded_space, spec->format.w); - if (spec->add_space) - ds_put_char (&trns->line, encoded_space); + const union value *input = ; + int x1; + } else { - ds_put_substring (&trns->line, ds_ss (&spec->string)); - if (0 != strcmp (trns->encoding, LEGACY_NATIVE)) - { - size_t length = ds_length (&spec->string); - char *data = ss_data (ds_tail (&trns->line, length)); - char *s = recode_string (trns->encoding, LEGACY_NATIVE, data, length); - memcpy (data, s, length); - free (s); - } } } - flush_records (trns, trns->record_cnt + 1, &eject, &record); if (trns->writer != NULL && dfm_write_error (trns->writer)) return TRNS_ERROR; @@ -516,23 +562,16 @@ flush_records (struct print_trns *trns, int target_record, { *eject = false; if (trns->writer == NULL) - som_eject_page (); + text_item_submit (text_item_create (TEXT_ITEM_EJECT_PAGE, "")); else leader = '1'; } - line[0] = legacy_from_native (trns->encoding, leader); + line[0] = recode_byte (trns->encoding, C_ENCODING, leader); if (trns->writer == NULL) - tab_output_text (TAB_FIX | TAT_NOWRAP, &line[1]); + tab_output_text (TAB_FIX, &line[1]); else - { - if (!trns->include_prefix) - { - line++; - length--; - } - dfm_put_record (trns->writer, line, length); - } + dfm_put_record (trns->writer, line, length); ds_truncate (&trns->line, 1); }