Mercurial > emacs
changeset 20724:e34b99d443b1
(Fcall_process): Use raw-text instead of emacs-mule
when enable-multibyte-characters is nil. Adjusted for the change
of encode_coding and decode_coding.
author | Kenichi Handa <handa@m17n.org> |
---|---|
date | Thu, 22 Jan 1998 01:26:45 +0000 |
parents | a0df1e68fc76 |
children | bd592c9c3ef6 |
files | src/callproc.c |
diffstat | 1 files changed, 36 insertions(+), 29 deletions(-) [+] |
line wrap: on
line diff
--- a/src/callproc.c Thu Jan 22 01:26:45 1998 +0000 +++ b/src/callproc.c Thu Jan 22 01:26:45 1998 +0000 @@ -285,7 +285,7 @@ if (!NILP (Vcoding_system_for_read)) val = Vcoding_system_for_read; else if (NILP (current_buffer->enable_multibyte_characters)) - val = Qemacs_mule; + val = Qraw_text; else { if (!EQ (coding_systems, Qt)) @@ -447,16 +447,16 @@ int size = encoding_buffer_size (&argument_coding, XSTRING (args[i])->size_byte); unsigned char *dummy1 = (unsigned char *) alloca (size); - int produced, dummy; + int dummy; /* The Irix 4.0 compiler barfs if we eliminate dummy. */ new_argv[i - 3] = dummy1; - produced = encode_coding (&argument_coding, - XSTRING (args[i])->data, - new_argv[i - 3], - XSTRING (args[i])->size_byte, - size, &dummy); - new_argv[i - 3][produced] = 0; + encode_coding (&argument_coding, + XSTRING (args[i])->data, + new_argv[i - 3], + XSTRING (args[i])->size_byte, + size); + new_argv[i - 3][argument_coding.produced] = 0; } UNGCPRO; } @@ -666,6 +666,7 @@ register int nread; int first = 1; int total_read = 0; + int carryover = 0; while (1) { @@ -673,11 +674,10 @@ of the buffer size we have. But don't read less than 1024--save that for the next bufferful. */ - nread = process_coding.carryover_size; /* This value is initially 0. */ + nread = carryover; while (nread < bufsize - 1024) { - int this_read - = read (fd[0], bufptr + nread, bufsize - nread); + int this_read = read (fd[0], bufptr + nread, bufsize - nread); if (this_read < 0) goto give_up; @@ -691,13 +691,13 @@ give_up_1: /* Now NREAD is the total amount of data in the buffer. */ - if (nread == 0) + if (nread == carryover) /* Here, just tell decode_coding that we are processing the last block. We break the loop after decoding. */ - process_coding.last_block = 1; + process_coding.mode |= CODING_MODE_LAST_BLOCK; immediate_quit = 0; - total_read += nread; + total_read += nread - carryover; if (!NILP (buffer)) { @@ -705,19 +705,32 @@ insert (bufptr, nread); else { /* We have to decode the input. */ - int size = decoding_buffer_size (&process_coding, bufsize); + int size = decoding_buffer_size (&process_coding, nread); char *decoding_buf = get_conversion_buffer (size); - int dummy; - nread = decode_coding (&process_coding, bufptr, decoding_buf, - nread, size, &dummy); - if (nread > 0) - insert (decoding_buf, nread); + decode_coding (&process_coding, bufptr, decoding_buf, + nread, size); + if (process_coding.produced > 0) + insert (decoding_buf, process_coding.produced); + carryover = process_coding.produced - process_coding.consumed; + if (carryover > 0) + { + /* As CARRYOVER should not be that large, we had + better avoid overhead of bcopy. */ + char *p = bufptr + process_coding.consumed; + char *pend = p + carryover; + char *dst = bufptr; + + while (p < pend) *dst++ = *p++; + } } } - - if (process_coding.last_block) - break; + if (process_coding.mode & CODING_MODE_LAST_BLOCK) + { + if (carryover > 0) + insert (bufptr, carryover); + break; + } /* Make the buffer bigger as we continue to read more data, but not past 64k. */ @@ -727,12 +740,6 @@ bufptr = (char *) alloca (bufsize); } - if (!NILP (buffer) && process_coding.carryover_size > 0) - /* We have carryover in the last decoding. It should be - processed again after reading more data. */ - bcopy (process_coding.carryover, bufptr, - process_coding.carryover_size); - if (!NILP (display) && INTERACTIVE) { if (first)