+#include "stream.h"
+#include <stdio.h>
+#include <string.h>
+#include <glib.h>
+#include <glib/gstdio.h>
+
+#define min(x,y) ( (x > y)? y : x )
+
+/*
+ *
+ **************** WRITING FUNCTIONS ********************************************
+ *
+ */
+
+/* Internal function: change illegal (control) characters in a string to a
+placeholder character. Must free returned string afterwards. */
+static gchar *
+remove_latin1_control_characters(unsigned char *s, gssize len)
+{
+ gchar *retval = g_new0(gchar, len);
+ int i;
+ for(i = 0; i < len; i++)
+ if( (s[i] < 32 && s[i] != 10) || (s[i] >= 127 && s[i] <= 159) )
+ retval[i] = '?';
+ /* Our placeholder character is '?'; other options are possible,
+ like printing "0x7F" or something */
+ else
+ retval[i] = s[i];
+ return retval;
+}
+
+/* Internal function: convert a Latin-1 string to a UTF-8 string, replacing
+Latin-1 control characters by a placeholder first. The UTF-8 string must be
+freed afterwards. Returns NULL on error. */
+static gchar *
+convert_latin1_to_utf8(gchar *s, gssize len)
+{
+ GError *error = NULL;
+ gchar *utf8;
+ gchar *canonical = remove_latin1_control_characters( (unsigned char *)s,
+ len);
+ utf8 = g_convert(canonical, len, "UTF-8", "ISO-8859-1", NULL, NULL, &error);
+ g_free(canonical);
+
+ if(utf8 == NULL)
+ {
+ error_dialog(NULL, error, "Error during latin1->utf8 conversion: ");
+ return NULL;
+ }
+
+ return utf8;
+}
+
+/* Internal function: write a UTF-8 string to a window's text buffer. */
+static void
+write_utf8_to_window(winid_t win, gchar *s)
+{
+ GtkTextBuffer *buffer =
+ gtk_text_view_get_buffer( GTK_TEXT_VIEW(win->widget) );
+
+ GtkTextIter iter;
+ gtk_text_buffer_get_end_iter(buffer, &iter);
+ gtk_text_buffer_insert(buffer, &iter, s, -1);
+}
+
+/* Internal function: write a UTF-8 buffer with length to a stream. */
+static void
+write_buffer_to_stream(strid_t str, gchar *buf, glui32 len)
+{
+ switch(str->stream_type)
+ {
+ case STREAM_TYPE_WINDOW:
+ /* Each window type has a different way of printing to it */
+ switch(str->window->window_type)
+ {
+ /* Printing to these windows' streams does nothing */
+ case wintype_Blank:
+ case wintype_Pair:
+ case wintype_Graphics:
+ str->write_count += len;
+ break;
+ /* Text buffer window */
+ case wintype_TextBuffer:
+ {
+ gchar *utf8 = convert_latin1_to_utf8(buf, len);
+ write_utf8_to_window(str->window, utf8);
+ g_free(utf8);
+ }
+ str->write_count += len;
+ break;
+ default:
+ g_warning("%s: Writing to this kind of window unsupported.",
+ __func__);
+ }
+
+ /* Now write the same buffer to the window's echo stream */
+ if(str->window->echo_stream != NULL)
+ write_buffer_to_stream(str->window->echo_stream, buf, len);
+
+ break;
+
+ case STREAM_TYPE_MEMORY:
+ if(str->unicode && str->ubuffer)
+ {
+ int foo = 0;
+ while(str->mark < str->buflen && foo < len)
+ str->ubuffer[str->mark++] = (glui32)buf[foo++];
+ }
+ if(!str->unicode && str->buffer)
+ {
+ memmove(str->buffer + str->mark, buf,
+ min(len, str->buflen - str->mark));
+ }
+
+ str->write_count += len;
+ break;
+
+ case STREAM_TYPE_FILE:
+ if(str->binary)
+ {
+ if(str->unicode)
+ {
+ /* Convert to four-byte big-endian */
+ gchar *writebuffer = g_new0(gchar, len * 4);
+ int i;
+ for(i = 0; i < len; i++)
+ writebuffer[i * 4 + 3] = buf[i];
+ fwrite(writebuffer, sizeof(gchar), len * 4,
+ str->file_pointer);
+ }
+ else /* Regular file */
+ {
+ fwrite(buf, sizeof(gchar), len, str->file_pointer);
+ }
+ }
+ else /* Text mode is the same for Unicode and regular files */
+ {
+ gchar *utf8 = convert_latin1_to_utf8(buf, len);
+ g_fprintf(str->file_pointer, "%s", utf8);
+ g_free(utf8);
+ }
+
+ str->write_count += len;
+ break;
+ default:
+ g_warning("%s: Writing to this kind of stream unsupported.",
+ __func__);
+ }
+}
+
+/**
+ * glk_put_char_stream:
+ * @str: An output stream.
+ * @ch: A character in Latin-1 encoding.
+ *
+ * Prints one character @ch to the stream @str. It is illegal for @str to be
+ * %NULL, or an input-only stream.
+ */
+void
+glk_put_char_stream(strid_t str, unsigned char ch)
+{
+ g_return_if_fail(str != NULL);
+ g_return_if_fail(str->file_mode != filemode_Read);
+
+ write_buffer_to_stream(str, (gchar *)&ch, 1);
+}
+
+/**
+ * glk_put_string_stream:
+ * @str: An output stream.
+ * @s: A null-terminated string in Latin-1 encoding.
+ *
+ * Prints @s to the stream @str. It is illegal for @str to be %NULL, or an
+ * input-only stream.
+ */
+void
+glk_put_string_stream(strid_t str, char *s)
+{
+ g_return_if_fail(str != NULL);
+ g_return_if_fail(str->file_mode != filemode_Read);
+
+ write_buffer_to_stream(str, (gchar *)s, strlen(s));
+}
+
+/**
+ * glk_put_buffer_stream:
+ * @str: An output stream.
+ * @buf: An array of characters in Latin-1 encoding.
+ * @len: Length of @buf.
+ *
+ * Prints @buf to the stream @str. It is illegal for @str to be %NULL, or an
+ * input-only stream.
+ */
+void
+glk_put_buffer_stream(strid_t str, char *buf, glui32 len)
+{
+ g_return_if_fail(str != NULL);
+ g_return_if_fail(str->file_mode != filemode_Read);
+
+ write_buffer_to_stream(str, (gchar *)buf, len);
+}
+
+/*
+ *
+ **************** READING FUNCTIONS ********************************************
+ *
+ */
+
+/* Internal function: Read one big-endian four-byte character from file fp and
+return it as a Unicode code point, or -1 on EOF */
+static glsi32
+read_ucs4be_char_from_file(FILE *fp)
+{
+ unsigned char readbuffer[4];
+ if(fread(readbuffer, sizeof(unsigned char), 4, fp) < 4)
+ return -1; /* EOF */
+ return
+ readbuffer[0] << 24 |
+ readbuffer[1] << 16 |
+ readbuffer[2] << 8 |
+ readbuffer[3];
+}
+
+/* Internal function: Read one UTF-8 character, which may be more than one byte,
+from file fp and return it as a Unicode code point, or -1 on EOF */
+static glsi32
+read_utf8_char_from_file(FILE *fp)
+{
+ gchar readbuffer[4] = {0, 0, 0, 0}; /* Max UTF-8 width */
+ int foo;
+ gunichar charresult = (gunichar)-2;
+ for(foo = 0; foo < 4 && charresult == (gunichar)-2; foo++)
+ {
+ int ch = fgetc(fp);
+ if(ch == EOF)
+ return -1;
+ readbuffer[foo] = (gchar)ch;
+ charresult = g_utf8_get_char_validated(readbuffer, foo);
+ /* charresult is -1 if invalid, -2 if incomplete, and the unicode code
+ point otherwise */
+ }
+ /* Silently return unknown characters as 0xFFFD, Replacement Character */
+ if(charresult == (gunichar)-1 || charresult == (gunichar)-2)
+ return 0xFFFD;
+ return charresult;
+}
+
+/* Internal function: Tell whether this code point is a Unicode newline. The
+file pointer and eight-bit flag are included in case the newline is a CR
+(U+000D). If the next character is LF (U+000A) then it also belongs to the
+newline. */
+static gboolean
+is_unicode_newline(glsi32 ch, FILE *fp, gboolean utf8)
+{
+ if(ch == 0x0A || ch == 0x85 || ch == 0x0C || ch == 0x2028 || ch == 0x2029)
+ return TRUE;
+ if(ch == 0x0D) {
+ glsi32 ch2 = utf8? read_utf8_char_from_file(fp) :
+ read_ucs4be_char_from_file(fp);
+ if(ch2 != 0x0A)
+ fseek(fp, utf8? -1 : -4, SEEK_CUR);
+ return TRUE;
+ }
+ return FALSE;
+}
+
+/**
+ * glk_get_char_stream:
+ * @str: An input stream.
+ *
+ * Reads one character from the stream @str. (There is no notion of a ``current
+ * input stream.'') It is illegal for @str to be %NULL, or an output-only
+ * stream.
+ *
+ * The result will be between 0 and 255. As with all basic text functions, Glk
+ * assumes the Latin-1 encoding. If the end of the stream has been reached, the
+ * result will be -1. Note that high-bit characters (128..255) are
+ * <emphasis>not</emphasis> returned as negative numbers.
+ *
+ * If the stream contains Unicode data --- for example, if it was created with
+ * glk_stream_open_file_uni() or glk_stream_open_memory_uni() --- then
+ * characters beyond 255 will be returned as 0x3F ("?").
+ *
+ * Returns: A character value between 0 and 255, or -1 on end of stream.
+ */
+glsi32
+glk_get_char_stream(strid_t str)
+{
+ g_return_val_if_fail(str != NULL, -1);
+ g_return_val_if_fail(str->file_mode == filemode_Read
+ || str->file_mode == filemode_ReadWrite, -1);
+
+ switch(str->stream_type)
+ {
+ case STREAM_TYPE_MEMORY:
+ if(str->unicode)
+ {
+ if(!str->ubuffer || str->mark >= str->buflen)
+ return -1;
+ glui32 ch = str->ubuffer[str->mark++];
+ str->read_count++;
+ return (ch > 0xFF)? 0x3F : ch;
+ }
+ else
+ {
+ if(!str->buffer || str->mark >= str->buflen)
+ return -1;
+ char ch = str->buffer[str->mark++];
+ str->read_count++;
+ return ch;
+ }
+ break;
+
+ case STREAM_TYPE_FILE:
+ if(str->binary)
+ {
+ if(str->unicode)
+ {
+ glsi32 ch = read_ucs4be_char_from_file(str->file_pointer);
+ if(ch == -1)
+ return -1;
+ str->read_count++;
+ return (ch > 0xFF)? 0x3F : ch;
+ }
+ else /* Regular file */
+ {
+ int ch = fgetc(str->file_pointer);
+ if(ch == EOF)
+ return -1;
+
+ str->read_count++;
+ return ch;
+ }
+ }
+ else /* Text mode is the same for Unicode and regular files */
+ {
+ glsi32 ch = read_utf8_char_from_file(str->file_pointer);
+ if(ch == -1)
+ return -1;
+
+ str->read_count++;
+ return (ch > 0xFF)? 0x3F : ch;
+ }
+ default:
+ g_warning("%s: Reading from this kind of stream unsupported.",
+ __func__);
+ return -1;
+ }
+}
+
+/**
+ * glk_get_buffer_stream:
+ * @str: An input stream.
+ * @buf: A buffer with space for at least @len characters.
+ * @len: The number of characters to read.
+ *
+ * Reads @len characters from @str, unless the end of stream is reached first.
+ * No terminal null is placed in the buffer.
+ *
+ * Returns: The number of characters actually read.
+ */
+glui32
+glk_get_buffer_stream(strid_t str, char *buf, glui32 len)
+{
+ g_return_val_if_fail(str != NULL, 0);
+ g_return_val_if_fail(str->file_mode == filemode_Read
+ || str->file_mode == filemode_ReadWrite, 0);
+ g_return_val_if_fail(buf != NULL, 0);
+
+ switch(str->stream_type)
+ {
+ case STREAM_TYPE_MEMORY:
+ {
+ int copycount = 0;
+ if(str->unicode)
+ {
+ while(copycount < len && str->ubuffer
+ && str->mark < str->buflen)
+ {
+ glui32 ch = str->ubuffer[str->mark++];
+ buf[copycount++] = (ch > 0xFF)? '?' : (char)ch;
+ }
+ }
+ else
+ {
+ if(str->buffer) /* if not, copycount stays 0 */
+ copycount = min(len, str->buflen - str->mark);
+ memmove(buf, str->buffer + str->mark, copycount);
+ }
+
+ str->read_count += copycount;
+ return copycount;
+ }
+ case STREAM_TYPE_FILE:
+ if(str->binary)
+ {
+ if(str->unicode) /* Binary file with 4-byte characters */
+ {
+ /* Read len characters of 4 bytes each */
+ unsigned char *readbuffer = g_new0(unsigned char, 4 * len);
+ size_t count = fread(readbuffer, sizeof(unsigned char),
+ 4 * len, str->file_pointer);
+ /* If there was an incomplete character */
+ if(count % 4 != 0)
+ {
+ count -= count % 4;
+ g_warning("%s: Incomplete character in binary Unicode "
+ "file.", __func__);
+ }
+
+ str->read_count += count / 4;
+ int foo;
+ for(foo = 0; foo < count; foo += 4)
+ {
+ glsi32 ch = readbuffer[foo] << 24
+ | readbuffer[foo + 1] << 16
+ | readbuffer[foo + 2] << 8
+ | readbuffer[foo + 3];
+ buf[foo / 4] = (ch > 255)? 0x3F : (char)ch;
+ }
+ g_free(readbuffer);
+ return count / 4;
+ }
+ else /* Regular binary file */
+ {
+ size_t count = fread(buf, sizeof(char), len,
+ str->file_pointer);
+ str->read_count += count;
+ return count;
+ }
+ }
+ else /* Text mode is the same for Unicode and regular files */
+ {
+ /* Do it character-by-character */
+ int foo;
+ for(foo = 0; foo < len; foo++)
+ {
+ glsi32 ch = read_utf8_char_from_file(str->file_pointer);
+ if(ch == -1)
+ break;
+ str->read_count++;
+ buf[foo] = (ch > 0xFF)? 0x3F : (gchar)ch;
+ }
+ return foo;
+ }
+ default:
+ g_warning("%s: Reading from this kind of stream unsupported.",
+ __func__);
+ return 0;
+ }
+}
+
+/**
+ * glk_get_line_stream:
+ * @str: An input stream.
+ * @buf: A buffer with space for at least @len characters.
+ * @len: The number of characters to read, plus one.
+ *
+ * Reads characters from @str, until either @len - 1 characters have been read
+ * or a newline has been read. It then puts a terminal null ('\0') aracter on
+ * the end. It returns the number of characters actually read, including the
+ * newline (if there is one) but not including the terminal null.
+ *
+ * It is usually more efficient to read several characters at once with
+ * glk_get_buffer_stream() or glk_get_line_stream(), as opposed to calling
+ * glk_get_char_stream() several times.
+ *
+ * Returns: The number of characters actually read.
+ */
+glui32
+glk_get_line_stream(strid_t str, char *buf, glui32 len)
+{
+ g_return_val_if_fail(str != NULL, 0);
+ g_return_val_if_fail(str->file_mode == filemode_Read
+ || str->file_mode == filemode_ReadWrite, 0);
+ g_return_val_if_fail(buf != NULL, 0);
+
+ switch(str->stream_type)
+ {
+ case STREAM_TYPE_MEMORY:
+ {
+ int copycount = 0;
+ if(str->unicode)
+ {
+ /* Do it character-by-character */
+ while(copycount < len - 1 && str->ubuffer
+ && str->mark < str->buflen)
+ {
+ glui32 ch = str->ubuffer[str->mark++];
+ /* Check for Unicode newline; slightly different than
+ in file streams */
+ if(ch == 0x0A || ch == 0x85 || ch == 0x0C || ch == 0x2028
+ || ch == 0x2029)
+ {
+ buf[copycount++] = '\n';
+ break;
+ }
+ if(ch == 0x0D)
+ {
+ if(str->ubuffer[str->mark] == 0x0A)
+ str->mark++; /* skip past next newline */
+ buf[copycount++] = '\n';
+ break;
+ }
+ buf[copycount++] = (ch > 0xFF)? '?' : (char)ch;
+ }
+ buf[copycount] = '\0';
+ }
+ else
+ {
+ if(str->buffer) /* if not, copycount stays 0 */
+ copycount = min(len, str->buflen - str->mark);
+ memccpy(buf, str->buffer + str->mark, '\n', copycount);
+ }
+
+ str->read_count += copycount;
+ return copycount;
+ }
+ case STREAM_TYPE_FILE:
+ if(str->binary)
+ {
+ if(str->unicode) /* Binary file with 4-byte characters */
+ {
+ /* Do it character-by-character */
+ int foo;
+ for(foo = 0; foo < len - 1; foo++)
+ {
+ glsi32 ch =
+ read_ucs4be_char_from_file(str->file_pointer);
+ if(ch == -1)
+ {
+ buf[foo] = '\0';
+ return foo - 1;
+ }
+ str->read_count++;
+ if(is_unicode_newline(ch, str->file_pointer, FALSE))
+ {
+ buf[foo] = '\n';
+ buf[foo + 1] = '\0';
+ return foo;
+ }
+ buf[foo] = (ch > 0xFF)? '?' : (char)ch;
+ }
+ buf[len] = '\0';
+ return foo;
+ }
+ else /* Regular binary file */
+ {
+ fgets(buf, len, str->file_pointer);
+ str->read_count += strlen(buf);
+ return strlen(buf);
+ }
+ }
+ else /* Text mode is the same for Unicode and regular files */
+ {
+ /* Do it character-by-character */
+ int foo;
+ for(foo = 0; foo < len - 1; foo++)
+ {
+ glsi32 ch = read_utf8_char_from_file(str->file_pointer);
+ if(ch == -1)
+ {
+ buf[foo] = '\0';
+ return foo - 1;
+ }
+ str->read_count++;
+ if(is_unicode_newline(ch, str->file_pointer, TRUE))
+ {
+ buf[foo] = '\n';
+ buf[foo + 1] = '\0';
+ return foo;
+ }
+ buf[foo] = (ch > 0xFF)? 0x3F : (char)ch;
+ }
+ buf[len] = '\0';
+ return foo;
+ }
+ default:
+ g_warning("%s: Reading from this kind of stream unsupported.",
+ __func__);
+ return 0;
+ }
+}
+
+/*
+ *
+ **************** SEEKING FUNCTIONS ********************************************
+ *
+ */
+
+/**
+ * glk_stream_get_position:
+ * @str: A file or memory stream.
+ *
+ * Returns the position of the read/write mark in @str. For memory streams and
+ * binary file streams, this is exactly the number of characters read or written
+ * from the beginning of the stream (unless you have moved the mark with
+ * glk_stream_set_position().) For text file streams, matters are more
+ * ambiguous, since (for example) writing one byte to a text file may store more
+ * than one character in the platform's native encoding. You can only be sure
+ * that the position increases as you read or write to the file.
+ *
+ * Additional complication: for Latin-1 memory and file streams, a character is
+ * a byte. For Unicode memory and file streams (those created by
+ * glk_stream_open_file_uni() and glk_stream_open_memory_uni()), a character is
+ * a 32-bit word. So in a binary Unicode file, positions are multiples of four
+ * bytes.
+ *
+ * Returns: position of the read/write mark in @str.
+ */
+glui32
+glk_stream_get_position(strid_t str)
+{
+ g_return_val_if_fail(str != NULL, 0);
+
+ switch(str->stream_type)
+ {
+ case STREAM_TYPE_MEMORY:
+ return str->mark;
+ case STREAM_TYPE_FILE:
+ return ftell(str->file_pointer);
+ default:
+ g_warning("%s: Seeking not supported on this type of stream.",
+ __func__);
+ return 0;
+ }
+}
+
+/**
+ * glk_stream_set_position:
+ * @str: A file or memory stream.
+ * @pos: The position to set the mark to, relative to @seekmode.
+ * @seekmode: One of #seekmode_Start, #seekmode_Current, or #seekmode_End.
+ *
+ * Sets the position of the read/write mark in @str. The position is controlled
+ * by @pos, and the meaning of @pos is controlled by @seekmode:
+ * <itemizedlist>
+ * <listitem>#seekmode_Start: @pos characters after the beginning of the file.
+ * </listitem>
+ * <listitem>#seekmode_Current: @pos characters after the current position
+ * (moving backwards if @pos is negative.)</listitem>
+ * <listitem>#seekmode_End: @pos characters after the end of the file. (@pos
+ * should always be zero or negative, so that this will move backwards to a
+ * position within the file.</listitem>
+ * </itemizedlist>
+ * It is illegal to specify a position before the beginning or after the end of
+ * the file.
+ *
+ * In binary files, the mark position is exact --- it corresponds with the
+ * number of characters you have read or written. In text files, this mapping
+ * can vary, because of linefeed conventions or other character-set
+ * approximations. glk_stream_set_position() and glk_stream_get_position()
+ * measure positions in the platform's native encoding --- after character
+ * cookery. Therefore, in a text stream, it is safest to use
+ * glk_stream_set_position() only to move to the beginning or end of a file, or
+ * to a position determined by glk_stream_get_position().
+ *
+ * Again, in Latin-1 streams, characters are bytes. In Unicode streams,
+ * characters are 32-bit words, or four bytes each.
+ */
+void
+glk_stream_set_position(strid_t str, glsi32 pos, glui32 seekmode)
+{
+ g_return_if_fail(str != NULL);
+ g_return_if_fail(!(seekmode == seekmode_Start && pos < 0));
+ g_return_if_fail(!(seekmode == seekmode_End || pos > 0));
+
+ switch(str->stream_type)
+ {
+ case STREAM_TYPE_MEMORY:
+ switch(seekmode)
+ {
+ case seekmode_Start: str->mark = pos; break;
+ case seekmode_Current: str->mark += pos; break;
+ case seekmode_End: str->mark = str->buflen + pos; break;
+ default:
+ g_assert_not_reached();
+ return;
+ }
+ break;
+ case STREAM_TYPE_FILE:
+ {
+ int whence;
+ switch(seekmode)
+ {
+ case seekmode_Start: whence = SEEK_SET; break;
+ case seekmode_Current: whence = SEEK_CUR; break;
+ case seekmode_End: whence = SEEK_END; break;
+ default:
+ g_assert_not_reached();
+ return;
+ }
+ fseek(str->file_pointer, pos, whence);
+ break;
+ }
+ default:
+ g_warning("%s: Seeking not supported on this type of stream.",
+ __func__);
+ return;
+ }
+}
+