Changeset 2725 in MondoRescue for branches/2.2.9/mindi-busybox/libbb/get_line_from_file.c

Timestamp:

Feb 25, 2011, 9:26:54 PM (13 years ago)

Author:

Bruno Cornec

Message:

• Update mindi-busybox to 1.18.3 to avoid problems with the tar command which is now failing on recent versions with busybox 1.7.3

File:

• branches/2.2.9/mindi-busybox/libbb/get_line_from_file.c (modified) (7 diffs)

branches/2.2.9/mindi-busybox/libbb/get_line_from_file.c

@@ -7 +7 @@
  * Copyright (C) 2001 Matt Krai
  *
- * Licensed under GPLv2 or later, see file LICENSE in this tarball for details.
+ * Licensed under GPLv2 or later, see file LICENSE in this source tree.
  */
 
@@ -13 +13 @@
 
 /* This function reads an entire line from a text file, up to a newline
- * or NUL byte, inclusive.  It returns a malloc'ed char * which must be
- * stored and free'ed by the caller.  If end is NULL '\n' isn't considered
- * end of line.  If end isn't NULL, length of the chunk read is stored in it.
- * Return NULL if EOF/error */
-
-char *bb_get_chunk_from_file(FILE *file, int *end)
+ * or NUL byte, inclusive.  It returns a malloc'ed char * which
+ * must be free'ed by the caller.  If end is NULL '\n' isn't considered
+ * end of line.  If end isn't NULL, length of the chunk is stored in it.
+ * If lineno is not NULL, *lineno is incremented for each line,
+ * and also trailing '\' is recognized as line continuation.
+ *
+ * Returns NULL if EOF/error. */
+char* FAST_FUNC bb_get_chunk_with_continuation(FILE *file, int *end, int *lineno)
 {
     int ch;
@@ -28 +30 @@
         /* grow the line buffer as necessary */
         if (idx >= linebufsz) {
-            linebufsz += 80;
+            linebufsz += 256;
             linebuf = xrealloc(linebuf, linebufsz);
         }
         linebuf[idx++] = (char) ch;
-        if (!ch || (end && ch == '\n'))
+        if (!ch)
             break;
+        if (end && ch == '\n') {
+            if (lineno == NULL)
+                break;
+            (*lineno)++;
+            if (idx < 2 || linebuf[idx-2] != '\\')
+                break;
+            idx -= 2;
+        }
     }
     if (end)
@@ -44 +54 @@
         //  return NULL;
         //}
-        linebuf = xrealloc(linebuf, idx+1);
+        linebuf = xrealloc(linebuf, idx + 1);
         linebuf[idx] = '\0';
     }
@@ -50 +60 @@
 }
 
+char* FAST_FUNC bb_get_chunk_from_file(FILE *file, int *end)
+{
+    return bb_get_chunk_with_continuation(file, end, NULL);
+}
+
 /* Get line, including trailing \n if any */
-char *xmalloc_fgets(FILE *file)
+char* FAST_FUNC xmalloc_fgets(FILE *file)
 {
     int i;
@@ -57 +72 @@
     return bb_get_chunk_from_file(file, &i);
 }
-
 /* Get line.  Remove trailing \n */
-char *xmalloc_getline(FILE *file)
+char* FAST_FUNC xmalloc_fgetline(FILE *file)
 {
     int i;
@@ -69 +83 @@
     return c;
 }
+
+#if 0
+/* GNUism getline() should be faster (not tested) than a loop with fgetc */
+
+/* Get line, including trailing \n if any */
+char* FAST_FUNC xmalloc_fgets(FILE *file)
+{
+    char *res_buf = NULL;
+    size_t res_sz;
+
+    if (getline(&res_buf, &res_sz, file) == -1) {
+        free(res_buf); /* uclibc allocates a buffer even on EOF. WTF? */
+        res_buf = NULL;
+    }
+//TODO: trimming to res_sz?
+    return res_buf;
+}
+/* Get line.  Remove trailing \n */
+char* FAST_FUNC xmalloc_fgetline(FILE *file)
+{
+    char *res_buf = NULL;
+    size_t res_sz;
+
+    res_sz = getline(&res_buf, &res_sz, file);
+
+    if ((ssize_t)res_sz != -1) {
+        if (res_buf[res_sz - 1] == '\n')
+            res_buf[--res_sz] = '\0';
+//TODO: trimming to res_sz?
+    } else {
+        free(res_buf); /* uclibc allocates a buffer even on EOF. WTF? */
+        res_buf = NULL;
+    }
+    return res_buf;
+}
+
+#endif
+
+#if 0
+/* Faster routines (~twice as fast). +170 bytes. Unused as of 2008-07.
+ *
+ * NB: they stop at NUL byte too.
+ * Performance is important here. Think "grep 50gigabyte_file"...
+ * Ironically, grep can't use it because of NUL issue.
+ * We sorely need C lib to provide fgets which reports size!
+ *
+ * Update:
+ * Actually, uclibc and glibc have it. man getline. It's GNUism,
+ *   but very useful one (if it's as fast as this code).
+ * TODO:
+ * - currently, sed and sort use bb_get_chunk_from_file and heavily
+ *   depend on its "stop on \n or \0" behavior, and STILL they fail
+ *   to handle all cases with embedded NULs correctly. So:
+ * - audit sed and sort; convert them to getline FIRST.
+ * - THEN ditch bb_get_chunk_from_file, replace it with getline.
+ * - provide getline implementation for non-GNU systems.
+ */
+
+static char* xmalloc_fgets_internal(FILE *file, int *sizep)
+{
+    int len;
+    int idx = 0;
+    char *linebuf = NULL;
+
+    while (1) {
+        char *r;
+
+        linebuf = xrealloc(linebuf, idx + 0x100);
+        r = fgets(&linebuf[idx], 0x100, file);
+        if (!r) {
+            /* need to terminate in case this is error
+             * (EOF puts NUL itself) */
+            linebuf[idx] = '\0';
+            break;
+        }
+        /* stupid. fgets knows the len, it should report it somehow */
+        len = strlen(&linebuf[idx]);
+        idx += len;
+        if (len != 0xff || linebuf[idx - 1] == '\n')
+            break;
+    }
+    *sizep = idx;
+    if (idx) {
+        /* xrealloc(linebuf, idx + 1) is up to caller */
+        return linebuf;
+    }
+    free(linebuf);
+    return NULL;
+}
+
+/* Get line, remove trailing \n */
+char* FAST_FUNC xmalloc_fgetline_fast(FILE *file)
+{
+    int sz;
+    char *r = xmalloc_fgets_internal(file, &sz);
+    if (r && r[sz - 1] == '\n')
+        r[--sz] = '\0';
+    return r; /* not xrealloc(r, sz + 1)! */
+}
+
+char* FAST_FUNC xmalloc_fgets(FILE *file)
+{
+    int sz;
+    return xmalloc_fgets_internal(file, &sz);
+}
+
+/* Get line, remove trailing \n */
+char* FAST_FUNC xmalloc_fgetline(FILE *file)
+{
+    int sz;
+    char *r = xmalloc_fgets_internal(file, &sz);
+    if (!r)
+        return r;
+    if (r[sz - 1] == '\n')
+        r[--sz] = '\0';
+    return xrealloc(r, sz + 1);
+}
+#endif