[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[PATCH 15/18] u8-istream: New library for reading a text file and recodi
From: |
Ben Pfaff |
Subject: |
[PATCH 15/18] u8-istream: New library for reading a text file and recoding to UTF-8. |
Date: |
Sat, 19 Mar 2011 17:10:01 -0700 |
This new library will be used in an upcoming commit.
---
src/libpspp/automake.mk | 2 +
src/libpspp/u8-istream.c | 475 +++++++++++++++++++++++++++++++++++++++
src/libpspp/u8-istream.h | 45 ++++
tests/automake.mk | 5 +
tests/libpspp/u8-istream-test.c | 126 +++++++++++
tests/libpspp/u8-istream.at | 142 ++++++++++++
6 files changed, 795 insertions(+), 0 deletions(-)
create mode 100644 src/libpspp/u8-istream.c
create mode 100644 src/libpspp/u8-istream.h
create mode 100644 tests/libpspp/u8-istream-test.c
create mode 100644 tests/libpspp/u8-istream.at
diff --git a/src/libpspp/automake.mk b/src/libpspp/automake.mk
index 5cf660a..a7c9283 100644
--- a/src/libpspp/automake.mk
+++ b/src/libpspp/automake.mk
@@ -88,6 +88,8 @@ src_libpspp_libpspp_la_SOURCES = \
src/libpspp/temp-file.h \
src/libpspp/tower.c \
src/libpspp/tower.h \
+ src/libpspp/u8-istream.c \
+ src/libpspp/u8-istream.h \
src/libpspp/version.h \
src/libpspp/zip-writer.c \
src/libpspp/zip-writer.h
diff --git a/src/libpspp/u8-istream.c b/src/libpspp/u8-istream.c
new file mode 100644
index 0000000..6d4d770
--- /dev/null
+++ b/src/libpspp/u8-istream.c
@@ -0,0 +1,475 @@
+/* PSPP - a program for statistical analysis.
+ Copyright (C) 2010, 2011 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
+ the Free Software Foundation, either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU General Public License for more details.
+
+ You should have received a copy of the GNU General Public License
+ along with this program. If not, see <http://www.gnu.org/licenses/>. */
+
+#include <config.h>
+
+#include "u8-istream.h"
+
+#include <assert.h>
+#include <errno.h>
+#include <fcntl.h>
+#include <iconv.h>
+#include <stdint.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <unistd.h>
+#include <unistr.h>
+
+#include "libpspp/assertion.h"
+#include "libpspp/cast.h"
+#include "libpspp/compiler.h"
+#include "libpspp/encoding-guesser.h"
+
+#include "gl/c-strcase.h"
+#include "gl/localcharset.h"
+#include "gl/minmax.h"
+
+enum u8_istream_state
+ {
+ S_AUTO, /* Stream encoding not yet known. */
+ S_UTF8, /* Stream encoding is known to be UTF-8. */
+ S_CONVERT /* Stream encoding is known but not UTF-8. */
+ };
+
+struct u8_istream
+ {
+ int fd;
+ iconv_t converter;
+ enum u8_istream_state state;
+
+ char *buffer;
+ char *head;
+ size_t length;
+
+ char outbuf[4];
+ size_t outlen;
+ };
+
+static ssize_t fill_buffer (struct u8_istream *);
+
+/* Opens FILENAME, which is encoded in FROMCODE, for reading as an UTF-8
+ stream, passing FLAGS to the open() function. Returns a new u8_istream if
+ successful, otherwise returns NULL and sets errno to an appropriate value.
+
+ The accepted forms for FROMCODE are listed at the top of
+ encoding-guesser.h. */
+struct u8_istream *
+u8_istream_for_file (const char *fromcode, const char *filename, int flags)
+{
+ struct u8_istream *is;
+ int fd;
+
+ assert (!(flags & O_CREAT));
+
+ fd = open (filename, flags);
+ if (fd < 0)
+ return NULL;
+
+ is = u8_istream_for_fd (fromcode, fd);
+ if (is == NULL)
+ {
+ int save_errno = errno;
+ close (fd);
+ errno = save_errno;
+ }
+
+ return is;
+}
+
+/* Creates and returns a new u8_istream that reads its input from FD. Returns
+ a new u8_istream if successful, otherwise returns NULL and sets errno to an
+ appropriate value.
+
+ The accepted forms for FROMCODE are listed at the top of
+ encoding-guesser.h. */
+struct u8_istream *
+u8_istream_for_fd (const char *fromcode, int fd)
+{
+ struct u8_istream *is;
+ const char *encoding;
+
+ is = malloc (sizeof *is);
+ if (is == NULL)
+ return NULL;
+
+ is->fd = fd;
+ is->converter = (iconv_t) -1;
+ is->buffer = malloc (U8_ISTREAM_BUFFER_SIZE);
+ if (is->buffer == NULL)
+ goto error;
+ is->head = is->buffer;
+ is->length = 0;
+ is->outlen = 0;
+
+ if (fill_buffer (is) < 0)
+ goto error;
+
+ encoding = encoding_guess_head_encoding (fromcode, is->buffer, is->length);
+ if (!strcmp (encoding, "UTF-8"))
+ is->state = S_UTF8;
+ else
+ {
+ if (encoding_guess_encoding_is_auto (fromcode)
+ && !strcmp (encoding, "ASCII"))
+ is->state = S_AUTO;
+ else
+ is->state = S_CONVERT;
+
+ is->converter = iconv_open ("UTF-8",
+ encoding_guess_parse_encoding (fromcode));
+ if (is->converter == (iconv_t) -1)
+ goto error;
+ }
+
+ return is;
+
+error:
+ u8_istream_free (is);
+ return NULL;
+}
+
+/* Closes IS and its underlying file descriptor and frees all associated
+ resources. Returns the return value from close(). */
+int
+u8_istream_close (struct u8_istream *is)
+{
+ if (is != NULL)
+ {
+ int fd = is->fd;
+ u8_istream_free (is);
+ return close (fd);
+ }
+ return 0;
+}
+
+/* Frees IS and associated resources, but does not close the underlying file
+ descriptor. (Thus, the client must close the file descriptor when it is no
+ longer needed.) */
+void
+u8_istream_free (struct u8_istream *is)
+{
+ if (is != NULL)
+ {
+ if (is->converter != (iconv_t) -1)
+ iconv_close (is->converter);
+ free (is->buffer);
+ free (is);
+ }
+}
+
+static void
+substitute_invalid_input_byte (struct u8_istream *is)
+{
+ assert (is->outlen == 0);
+ is->head++;
+ is->length--;
+ is->outlen = u8_uctomb (CHAR_CAST (uint8_t *, is->outbuf),
+ 0xfffd, sizeof is->outbuf);
+}
+
+static ssize_t
+fill_buffer (struct u8_istream *is)
+{
+ ssize_t n;
+
+ /* Move any unused bytes to the beginning of the input buffer. */
+ if (is->length > 0 && is->buffer != is->head)
+ memmove (is->buffer, is->head, is->length);
+ is->head = is->buffer;
+
+ /* Read more input. */
+ n = read (is->fd, is->buffer + is->length,
+ U8_ISTREAM_BUFFER_SIZE - is->length);
+ if (n > 0)
+ is->length += n;
+ return n;
+}
+
+static ssize_t
+read_auto (struct u8_istream *is, char *buffer, size_t size)
+{
+ size_t original_size = size;
+ int retval = 0;
+
+ while (size > 0)
+ {
+ if (is->length > 0)
+ {
+ size_t n_ascii;
+
+ n_ascii = encoding_guess_count_ascii (is->head,
+ MIN (is->length, size));
+
+ memcpy (buffer, is->head, n_ascii);
+ buffer += n_ascii;
+ size -= n_ascii;
+
+ is->head += n_ascii;
+ is->length -= n_ascii;
+
+ if (size == 0)
+ break;
+ }
+
+ if (is->length == 0)
+ {
+ retval = fill_buffer (is);
+ if (retval > 0)
+ continue;
+ else
+ break;
+ }
+
+ /* is->head points to a byte that isn't a printable ASCII character.
+ Fill up the buffer and check for UTF-8. */
+ fill_buffer (is);
+ is->state = (encoding_guess_tail_is_utf8 (is->head, is->length)
+ ? S_UTF8 : S_CONVERT);
+ if (size == original_size)
+ return u8_istream_read (is, buffer, size);
+ break;
+ }
+
+ return original_size - size;
+}
+
+static int
+convert_iconv (iconv_t converter,
+ char **inbufp, size_t *inbytesleft,
+ char **outbufp, size_t *outbytesleft)
+{
+ size_t n = iconv (converter, inbufp, inbytesleft, outbufp, outbytesleft);
+ return n == SIZE_MAX ? errno : 0;
+}
+
+static int
+convert_utf8 (iconv_t converter UNUSED,
+ char **inbufp, size_t *inbytesleft,
+ char **outbufp, size_t *outbytesleft)
+{
+ const uint8_t *in = CHAR_CAST (const uint8_t *, *inbufp);
+ size_t n = MIN (*inbytesleft, *outbytesleft);
+ size_t ofs = 0;
+ int error;
+
+ for (;;)
+ {
+ ucs4_t uc;
+ int mblen;
+
+ if (ofs >= n)
+ {
+ error = ofs < *inbytesleft ? E2BIG : 0;
+ break;
+ }
+
+ mblen = u8_mbtouc (&uc, in + ofs, n - ofs);
+ if (uc == 0xfffd)
+ {
+ int retval = u8_mbtoucr (&uc, in + ofs, *inbytesleft - ofs);
+ if (retval == mblen)
+ {
+ /* There's an actual U+FFFD in the input stream. Carry on. */
+ }
+ else
+ {
+ error = (retval == -1 ? EILSEQ
+ : retval == -2 ? EINVAL
+ : E2BIG);
+ break;
+ }
+ }
+
+ ofs += mblen;
+ }
+
+ if (ofs > 0)
+ {
+ memcpy (*outbufp, *inbufp, ofs);
+ *inbufp += ofs;
+ *inbytesleft -= ofs;
+ *outbufp += ofs;
+ *outbytesleft -= ofs;
+ }
+
+ return error;
+}
+
+static ssize_t
+read_convert (struct u8_istream *is,
+ int (*convert) (iconv_t converter,
+ char **inbufp, size_t *inbytesleft,
+ char **outbufp, size_t *outbytesleft),
+ char *buffer, size_t size)
+{
+ size_t original_size = size;
+
+ while (size > 0)
+ {
+ ssize_t n_read;
+
+ if (is->outlen > 0)
+ {
+ size_t n = MIN (size, is->outlen);
+
+ memcpy (buffer, is->outbuf, n);
+ is->outlen -= n;
+ if (is->outlen > 0)
+ memmove (is->outbuf, is->outbuf + n, is->outlen);
+
+ buffer += n;
+ size -= n;
+
+ if (size == 0)
+ break;
+ }
+
+ if (is->length)
+ {
+ int error = convert (is->converter,
+ &is->head, &is->length,
+ &buffer, &size);
+ if (size == 0)
+ break;
+
+ switch (error)
+ {
+ case 0:
+ /* Converted all of the input into output, possibly with space
+ for output left over.
+
+ Read more input. */
+ break;
+
+ case EILSEQ:
+ substitute_invalid_input_byte (is);
+ continue;
+
+ case EINVAL:
+ /* Incomplete byte sequence at end of input. Read more
+ input. */
+ break;
+
+ default:
+ /* A real error of some kind (ENOMEM?). */
+ return -1;
+
+ case E2BIG:
+ /* Ran out of room for output.
+ Convert into outbuf and copy from there instead. */
+ {
+ char *outptr = is->outbuf;
+ size_t outleft = sizeof is->outbuf;
+
+ error = convert (is->converter,
+ &is->head, &is->length,
+ &outptr, &outleft);
+ is->outlen = outptr - is->outbuf;
+ if (is->outlen > 0)
+ continue;
+
+ switch (error)
+ {
+ case EILSEQ:
+ substitute_invalid_input_byte (is);
+ continue;
+
+ case E2BIG:
+ case EINVAL:
+ continue;
+
+ default:
+ /* A real error of some kind (ENOMEM?). */
+ return -1;
+ }
+ }
+ }
+ }
+
+ assert (is->length <= MB_LEN_MAX);
+ n_read = fill_buffer (is);
+ if (n_read <= 0)
+ {
+ if (original_size != size)
+ {
+ /* We produced some output so don't report EOF or error yet. */
+ break;
+ }
+ else if (n_read == 0 && is->length != 0)
+ {
+ /* Incomplete byte sequence at end of file. */
+ substitute_invalid_input_byte (is);
+ }
+ else
+ {
+ /* Propagate end-of-file or error to caller. */
+ return n_read;
+ }
+ }
+ }
+
+ return original_size - size;
+}
+
+/* Reads up to SIZE bytes of UTF-8 text from IS into BUFFER. Returns the
+ number of bytes read if successful, 0 at end of file, or -1 if an error
+ occurred before any data could be read. Upon error, sets errno to an
+ appropriate value. */
+ssize_t
+u8_istream_read (struct u8_istream *is, char *buffer, size_t size)
+{
+ switch (is->state)
+ {
+ case S_CONVERT:
+ return read_convert (is, convert_iconv, buffer, size);
+
+ case S_AUTO:
+ return read_auto (is, buffer, size);
+
+ case S_UTF8:
+ return read_convert (is, convert_utf8, buffer, size);
+ }
+
+ NOT_REACHED ();
+}
+
+/* Returns the file descriptor underlying IS. */
+int
+u8_istream_fileno (const struct u8_istream *is)
+{
+ return is->fd;
+}
+
+/* Test functions.
+
+ These functions are probably useful only for white-box testing. */
+
+/* Returns true if the encoding of the file being read by IS is not yet
+ known. */
+bool
+u8_istream_is_auto (const struct u8_istream *is)
+{
+ return is->state == S_AUTO;
+}
+
+/* Returns true if the encoding of the file being read by IS has been
+ determined to be UTF-8. */
+bool
+u8_istream_is_utf8 (const struct u8_istream *is)
+{
+ return is->state == S_UTF8;
+}
diff --git a/src/libpspp/u8-istream.h b/src/libpspp/u8-istream.h
new file mode 100644
index 0000000..3e2acee
--- /dev/null
+++ b/src/libpspp/u8-istream.h
@@ -0,0 +1,45 @@
+/* PSPP - a program for statistical analysis.
+ Copyright (C) 2010, 2011 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
+ the Free Software Foundation, either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU General Public License for more details.
+
+ You should have received a copy of the GNU General Public License
+ along with this program. If not, see <http://www.gnu.org/licenses/>. */
+
+#ifndef LIBPSPP_U8_ISTREAM_H
+#define LIBPSPP_U8_ISTREAM_H 1
+
+#include <sys/types.h>
+#include <stdbool.h>
+
+/* u8_istream.
+
+ Reads a text file and reencodes its contents into UTF-8, with optional
+ automatic encoding detection.
+*/
+
+#define U8_ISTREAM_BUFFER_SIZE 4096
+
+struct u8_istream *u8_istream_for_fd (const char *fromcode, int fd);
+struct u8_istream *u8_istream_for_file (const char *fromcode,
+ const char *filename, int flags);
+
+int u8_istream_close (struct u8_istream *);
+void u8_istream_free (struct u8_istream *);
+
+ssize_t u8_istream_read (struct u8_istream *, char *, size_t);
+
+int u8_istream_fileno (const struct u8_istream *);
+
+bool u8_istream_is_auto (const struct u8_istream *);
+bool u8_istream_is_utf8 (const struct u8_istream *);
+
+#endif /* libpspp/u8-istream.h */
diff --git a/tests/automake.mk b/tests/automake.mk
index 7ef7d42..0b4a825 100644
--- a/tests/automake.mk
+++ b/tests/automake.mk
@@ -24,6 +24,7 @@ check_PROGRAMS += \
tests/libpspp/string-set-test \
tests/libpspp/stringi-set-test \
tests/libpspp/tower-test \
+ tests/libpspp/u8-istream-test \
tests/output/render-test
check-programs: $(check_PROGRAMS)
@@ -161,6 +162,9 @@ tests_libpspp_tower_test_SOURCES = \
tests_libpspp_tower_test_LDADD = gl/libgl.la $(LIBINTL)
tests_libpspp_tower_test_CPPFLAGS = $(AM_CPPFLAGS) -DASSERT_LEVEL=10
+tests_libpspp_u8_istream_test_SOURCES = tests/libpspp/u8-istream-test.c
+tests_libpspp_u8_istream_test_LDADD = src/libpspp/libpspp.la gl/libgl.la
+
tests_libpspp_sparse_array_test_SOURCES = \
src/libpspp/sparse-array.c \
src/libpspp/pool.c \
@@ -336,6 +340,7 @@ TESTSUITE_AT = \
tests/libpspp/string-set.at \
tests/libpspp/stringi-set.at \
tests/libpspp/tower.at \
+ tests/libpspp/u8-istream.at \
tests/math/moments.at \
tests/math/randist.at \
tests/output/charts.at \
diff --git a/tests/libpspp/u8-istream-test.c b/tests/libpspp/u8-istream-test.c
new file mode 100644
index 0000000..ab1b717
--- /dev/null
+++ b/tests/libpspp/u8-istream-test.c
@@ -0,0 +1,126 @@
+/* PSPP - a program for statistical analysis.
+ Copyright (C) 2010, 2011 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
+ the Free Software Foundation, either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU General Public License for more details.
+
+ You should have received a copy of the GNU General Public License
+ along with this program. If not, see <http://www.gnu.org/licenses/>. */
+
+#include <config.h>
+
+#include "libpspp/u8-istream.h"
+
+#include <errno.h>
+#include <fcntl.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <unistd.h>
+
+#include "libpspp/i18n.h"
+
+#include "gl/error.h"
+#include "gl/progname.h"
+#include "gl/xalloc.h"
+
+static void
+usage (void)
+{
+ printf ("usage: %s COMMAND [ARG]...\n"
+ "The available commands are:\n"
+ " help\n"
+ " print this usage message\n"
+ " buffer-size\n"
+ " print the buffer size, in bytes, on stdout\n"
+ " read FILE ENCODING [OUTBUF]\n"
+ " read FILE encoded in ENCODING (with output buffer size\n"
+ " OUTBUF) and print it on stdout in UTF-8\n",
+ program_name);
+ exit (0);
+}
+
+static void
+cmd_read (int argc, char *argv[])
+{
+ struct u8_istream *is;
+ const char *encoding;
+ const char *filename;
+ int outbufsize;
+ char *buffer;
+
+ if (argc < 4 || argc > 5)
+ error (1, 0, "bad syntax for `%s' command; use `%s help' for help",
+ argv[1], program_name);
+
+ outbufsize = argc > 4 ? atoi (argv[4]) : 4096;
+ buffer = xmalloc (outbufsize);
+
+ filename = argv[2];
+ encoding = *argv[3] ? argv[3] : NULL;
+
+ is = (!strcmp(filename, "-")
+ ? u8_istream_for_fd (encoding, STDIN_FILENO)
+ : u8_istream_for_file (encoding, filename, O_RDONLY));
+ if (is == NULL)
+ error (1, errno, "u8_istream_open failed");
+
+ if (u8_istream_is_auto (is))
+ printf ("Auto mode\n");
+ else if (u8_istream_is_utf8 (is))
+ printf ("UTF-8 mode\n");
+
+ for (;;)
+ {
+ ssize_t n;
+
+ n = u8_istream_read (is, buffer, outbufsize);
+ if (n > 0)
+ fwrite (buffer, 1, n, stdout);
+ else if (n < 0)
+ error (1, errno, "u8_istream_read failed");
+ else
+ break;
+ }
+
+ if (u8_istream_is_auto (is))
+ printf ("Auto mode\n");
+ else if (u8_istream_is_utf8 (is))
+ printf ("UTF-8 mode\n");
+
+ if (!strcmp(filename, "-"))
+ u8_istream_free (is);
+ else
+ {
+ if (u8_istream_close (is) != 0)
+ error (1, errno, "u8_istream_close failed");
+ }
+}
+
+int
+main (int argc, char *argv[])
+{
+ set_program_name (argv[0]);
+ i18n_init ();
+
+ if (argc < 2)
+ error (1, 0, "missing command name; use `%s help' for help", program_name);
+ else if (!strcmp(argv[1], "help") || !strcmp(argv[1], "--help"))
+ usage ();
+ else if (!strcmp(argv[1], "buffer-size"))
+ printf ("%d\n", U8_ISTREAM_BUFFER_SIZE);
+ else if (!strcmp(argv[1], "read"))
+ cmd_read (argc, argv);
+ else
+ error (1, 0, "unknown command `%s'; use `%s help' for help",
+ argv[1], program_name);
+
+ return 0;
+}
diff --git a/tests/libpspp/u8-istream.at b/tests/libpspp/u8-istream.at
new file mode 100644
index 0000000..2d8baa4
--- /dev/null
+++ b/tests/libpspp/u8-istream.at
@@ -0,0 +1,142 @@
+AT_BANNER([u8_istream])
+
+AT_SETUP([read ASCII])
+AT_KEYWORDS([u8_istream])
+AT_CHECK([supports_encodings ASCII])
+AT_CHECK([echo string | u8-istream-test read - ASCII], [0], [string
+])
+AT_CLEANUP
+
+AT_SETUP([read UTF-8])
+AT_KEYWORDS([u8_istream])
+AT_CHECK([printf '\346\227\245\346\234\254\350\252\236\n' | u8-istream-test
read - UTF-8], [0], [dnl
+UTF-8 mode
+日本語
+UTF-8 mode
+])
+AT_CLEANUP
+
+AT_SETUP([read EUC-JP])
+AT_KEYWORDS([u8_istream])
+AT_CHECK([supports_encodings EUC-JP])
+AT_CHECK([printf '\244\241 \244\242 \244\243 \244\244 \244\245 \244\246
\244\247 \244\250 \244\251 \244\252\n' | u8-istream-test read - EUC-JP],
+ [0],
+ [ぁ あ ぃ い ぅ う ぇ え ぉ お
+])
+AT_CLEANUP
+
+AT_SETUP([read UTF-8 with character split across input buffers])
+AT_KEYWORDS([u8_istream])
+buffer_size=`u8-istream-test buffer-size`
+($PERL -e "print 'x' x ($buffer_size - 16)"
+ printf
'\343\201\201\343\201\202\343\201\203\343\201\204\343\201\205\343\201\206\343\201\207\343\201\210\343\201\211\343\201\212\n')
> input
+(echo "UTF-8 mode"
+ cat input
+ echo "UTF-8 mode") > expout
+AT_CHECK([u8-istream-test read input UTF-8 16], [0], [expout])
+AT_CLEANUP
+
+AT_SETUP([read UTF-8 with character split across output buffers])
+AT_KEYWORDS([u8_istream])
+AT_CHECK([printf
'\343\201\201\343\201\202\343\201\203\343\201\204\343\201\205\343\201\206\343\201\207\343\201\210\343\201\211\343\201\212\n'
| u8-istream-test read - UTF-8 16], [0], [dnl
+UTF-8 mode
+ぁあぃいぅうぇえぉお
+UTF-8 mode
+])
+AT_CLEANUP
+
+AT_SETUP([read UTF-8 with character split across input and output buffers])
+AT_KEYWORDS([u8_istream])
+buffer_size=`u8-istream-test buffer-size`
+($PERL -e "print 'x' x ($buffer_size - 16)"
+ printf
'\343\201\201\343\201\202\343\201\203\343\201\204\343\201\205\343\201\206\343\201\207\343\201\210\343\201\211\343\201\212\n')
> input
+(echo "UTF-8 mode"
+ cat input
+ echo "UTF-8 mode") > expout
+AT_CHECK([u8-istream-test read input UTF-8 16], [0], [expout])
+AT_CLEANUP
+
+AT_SETUP([read EUC-JP with character split across input buffers])
+AT_KEYWORDS([u8_istream])
+AT_CHECK([supports_encodings EUC-JP])
+buffer_size=`u8-istream-test buffer-size`
+($PERL -e "print 'x' x ($buffer_size - 16)"
+ printf '\244\241 \244\242 \244\243 \244\244 \244\245 \244\246 \244\247 '
+ printf '\244\250 \244\251 \244\252\n') > input
+($PERL -e "print 'x' x ($buffer_size - 16)"
+ printf '\343\201\201\040\343\201\202\040\343\201\203\040\343\201\204\040'
+ printf '\343\201\205\040\343\201\206\040\343\201\207\040\343\201\210\040'
+ printf '\343\201\211\040\343\201\212\n') > expout
+AT_CHECK([u8-istream-test read input EUC-JP], [0], [expout])
+AT_CLEANUP
+
+AT_SETUP([read EUC-JP with character split across output buffers])
+AT_KEYWORDS([u8_istream])
+AT_CHECK([supports_encodings EUC-JP])
+AT_CHECK([printf
'\244\241\244\242\244\243\244\244\244\245\244\246\244\247\244\250\244\251\244\252\n'
| u8-istream-test read - EUC-JP 16],
+ [0],
+ [ぁあぃいぅうぇえぉお
+])
+AT_CLEANUP
+
+AT_SETUP([read EUC-JP with character split across input and output buffers])
+AT_KEYWORDS([u8_istream])
+AT_CHECK([supports_encodings EUC-JP])
+buffer_size=`u8-istream-test buffer-size`
+($PERL -e "print 'x' x ($buffer_size - 16)"
+ printf 'xyz\244\241\244\242\244\243\244\244\244\245\244\246\244\247\244\250'
+ printf '\244\251\244\252\n') > input
+($PERL -e "print 'x' x ($buffer_size - 16)"
+ printf '\170\171\172\343\201\201\343\201\202\343\201\203\343\201\204\343'
+ printf '\201\205\343\201\206\343\201\207\343\201\210\343\201\211\343\201'
+ printf '\212\n') > expout
+AT_CHECK([u8-istream-test read input EUC-JP 16], [0], [expout])
+AT_CLEANUP
+
+AT_SETUP([read ASCII as Auto])
+AT_KEYWORDS([u8_istream])
+AT_CHECK([echo string | u8-istream-test read - Auto], [0], [dnl
+Auto mode
+string
+Auto mode
+])
+AT_CLEANUP
+
+AT_SETUP([read UTF-8 as Auto])
+AT_KEYWORDS([u8_istream])
+AT_CHECK([printf 'entr\303\251e\n' | u8-istream-test read - Auto], [0], [dnl
+Auto mode
+entrée
+UTF-8 mode
+])
+AT_CLEANUP
+
+AT_SETUP([read ISO-8859-1 as Auto,ISO-8859-1])
+AT_KEYWORDS([u8_istream])
+AT_CHECK([supports_encodings ISO-8859-1])
+buffer_size=`u8-istream-test buffer-size`
+($PERL -e "print 'x' x int($buffer_size * 2.5)"; printf 'entr\351e\n') > input
+(echo "Auto mode"
+ $PERL -e "print 'x' x int($buffer_size * 2.5)"
+ printf 'entr\303\251e\n') > expout
+AT_CHECK([u8-istream-test read input Auto,ISO-8859-1], [0], [expout])
+AT_CLEANUP
+
+dnl UTF-16BE is not ASCII compatible so this doesn't start out in Auto mode.
+AT_SETUP([read UTF-16BE as Auto,UTF-16BE])
+AT_KEYWORDS([u8_istream])
+AT_CHECK([supports_encodings UTF-16BE])
+AT_CHECK([printf '\0e\0n\0t\0r\0\351\0e\0\n' | u8-istream-test read -
Auto,UTF-16BE],
+ [0], [dnl
+entrée
+])
+AT_CLEANUP
+
+AT_SETUP([read EUC-JP as Auto,EUC-JP])
+AT_KEYWORDS([u8_istream])
+AT_CHECK([supports_encodings EUC-JP])
+AT_CHECK([printf 'entr\217\253\261e\n' | u8-istream-test read - Auto,EUC-JP],
+ [0], [entrée
+])
+AT_CLEANUP
+
--
1.7.2.3
- [PATCH 14/18] encoding-guesser: New library to guess the encoding of a text file., (continued)
- [PATCH 14/18] encoding-guesser: New library to guess the encoding of a text file., Ben Pfaff, 2011/03/19
- [PATCH 04/18] output: New function text_item_create_nocopy()., Ben Pfaff, 2011/03/19
- [PATCH 05/18] str: New function ss_realloc()., Ben Pfaff, 2011/03/19
- [PATCH 13/18] i18n: New functions and data structure for obtaining encoding info., Ben Pfaff, 2011/03/19
- [PATCH 06/18] str: Rename ss_chomp() to ss_chomp_byte(), ds_chomp() to ds_chomp_byte()., Ben Pfaff, 2011/03/19
- [PATCH 02/18] file-name: Do not make output files line-buffered in fn_open()., Ben Pfaff, 2011/03/19
- [PATCH 12/18] identifier: Rename token_type_to_string() and make a new version., Ben Pfaff, 2011/03/19
- [PATCH 08/18] hash-functions: New function hash_case_bytes()., Ben Pfaff, 2011/03/19
- [PATCH 11/18] i18n: New functions for truncating strings in an arbitrary encoding., Ben Pfaff, 2011/03/19
- [PATCH 17/18] scan: New library for high-level PSPP syntax lexical analysis., Ben Pfaff, 2011/03/19
- [PATCH 15/18] u8-istream: New library for reading a text file and recoding to UTF-8.,
Ben Pfaff <=
- [PATCH 16/18] segment: New library for low-level phase of lexical syntax analysis., Ben Pfaff, 2011/03/19
- [PATCH 03/18] sys-file-reader: Refactor to clean up character encoding support., Ben Pfaff, 2011/03/19
- Re: [PATCH 00/18] rewrite PSPP lexer, John Darrington, 2011/03/20
- Re: [PATCH 00/18] rewrite PSPP lexer, John Darrington, 2011/03/22