libbb: experimental faster string reading routines.

2 files changed, 75 insertions, 1 deletions

diff --git a/include/libbb.h b/include/libbb.h
index 671584b43..2bd614c71 100644
--- a/include/libbb.h
+++ b/include/libbb.h
@@ -607,6 +607,8 @@ extern void xopen_xwrite_close(const char* file, const char *str) FAST_FUNC;
 
 /* Reads and prints to stdout till eof, then closes FILE. Exits on error: */
 extern void xprint_and_close_file(FILE *file) FAST_FUNC;
+
+extern char *bb_get_chunk_from_file(FILE *file, int *end) FAST_FUNC;
 /* Reads up to (and including) TERMINATING_STRING: */
 extern char *xmalloc_fgets_str(FILE *file, const char *terminating_string) FAST_FUNC;
 /* Chops off TERMINATING_STRING from the end: */
@@ -615,7 +617,9 @@ extern char *xmalloc_fgetline_str(FILE *file, const char *terminating_string) FA
 extern char *xmalloc_fgets(FILE *file) FAST_FUNC;
 /* Chops off '\n' from the end, unlike fgets: */
 extern char *xmalloc_fgetline(FILE *file) FAST_FUNC;
-extern char *bb_get_chunk_from_file(FILE *file, int *end) FAST_FUNC;
+/* Same, but doesn't try to conserve space (may have some slack after the end) */
+extern char *xmalloc_fgetline_fast(FILE *file) FAST_FUNC;
+
 extern void die_if_ferror(FILE *file, const char *msg) FAST_FUNC;
 extern void die_if_ferror_stdout(void) FAST_FUNC;
 extern void xfflush_stdout(void) FAST_FUNC;
diff --git a/libbb/get_line_from_file.c b/libbb/get_line_from_file.c
index 3a76f49a0..7b65ced8d 100644
--- a/libbb/get_line_from_file.c
+++ b/libbb/get_line_from_file.c
@@ -67,3 +67,73 @@ char* FAST_FUNC xmalloc_fgetline(FILE *file)
 
 	return c;
 }
+
+/* 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"...
+ * Iironically, grep can't use it because of NUL issue.
+ * We sorely need C lib to provide fgets which reports size!
+ */
+
+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)! */
+}
+
+#if 0
+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