For API input field parsing, use an array of structs instead of an array of integers.

7 files changed, 69 insertions, 29 deletions

diff --git a/ChangeLog b/ChangeLog
index a5fa974c..1d211a02 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,22 @@
+2017-03-09         Andrew J. Schorr     <aschorr@telemetry-investments.com>
+
+	* gawkapi.h (awk_input_field_info_t): Define new structure to contain
+	API field parsing info.
+	(awk_input_buf_t): Update get_record prototype to use an array of
+	awk_input_field_info_t instead of integers.
+	* awk.h (set_record): Change 3rd argument from 'const int *' to
+	'const awk_input_field_info_t *'.
+	* field.c (api_fw): Now points to an array of awk_input_field_info_t
+	instead of integers.
+	(set_record): Change 3rd argument to point to an array of
+	awk_input_field_info_t.
+	(api_parse_field): Update parsing logic to use awk_input_field_info_t
+	structures instead of an array of integers.
+	* io.c (inrec, do_getline_redir, do_getline): Change field_width type
+	from 'const int *' to 'const awk_input_field_info_t *'.
+	(get_a_record): Change field_width argument type from 'const int **'
+	to 'const awk_input_field_info_t **'.
+
 2017-03-09         Arnold D. Robbins     <arnold@skeeve.com>
 
 	* field.c: Minor style edits.
diff --git a/awk.h b/awk.h
index be6a6479..7acb2714 100644
--- a/awk.h
+++ b/awk.h
@@ -1510,7 +1510,7 @@ extern NODE *get_actual_argument(NODE *, int, bool);
 #endif
 /* field.c */
 extern void init_fields(void);
-extern void set_record(const char *buf, int cnt, const int *);
+extern void set_record(const char *buf, int cnt, const awk_input_field_info_t *);
 extern void reset_record(void);
 extern void rebuild_record(void);
 extern void set_NF(void);
diff --git a/extension/ChangeLog b/extension/ChangeLog
index 5d9a194e..bb99f9d8 100644
--- a/extension/ChangeLog
+++ b/extension/ChangeLog
@@ -1,3 +1,10 @@
+2017-03-09         Andrew J. Schorr     <aschorr@telemetry-investments.com>
+
+	* readdir_test.c (open_directory_t): Update field_width type from an
+	array of integers to an array of awk_input_field_info_t.
+	(dir_get_record): Ditto.
+	(dir_take_control_of): Ditto.
+
 2017-03-07         Andrew J. Schorr     <aschorr@telemetry-investments.com>
 
 	* Makefile.am (pkgextension_LTLIBRARIES): Remove testext.la, since it
diff --git a/extension/readdir_test.c b/extension/readdir_test.c
index cc23081e..e023b67c 100644
--- a/extension/readdir_test.c
+++ b/extension/readdir_test.c
@@ -85,7 +85,7 @@ int plugin_is_GPL_compatible;
 typedef struct open_directory {
 	DIR *dp;
 	char *buf;
-	int field_width[7];
+	awk_input_field_info_t field_width[4];
 } open_directory_t;
 
 /* ftype --- return type of file as a single character string */
@@ -169,7 +169,8 @@ get_inode(struct dirent *entry, const char *dirname)
 
 static int
 dir_get_record(char **out, awk_input_buf_t *iobuf, int *errcode,
-		char **rt_start, size_t *rt_len, const int **field_width)
+		char **rt_start, size_t *rt_len,
+		const awk_input_field_info_t **field_width)
 {
 	DIR *dp;
 	struct dirent *dirent;
@@ -206,13 +207,13 @@ dir_get_record(char **out, awk_input_buf_t *iobuf, int *errcode,
 #else
 	len = sprintf(the_dir->buf, "%llu", ino);
 #endif
-	the_dir->field_width[1] = len;
+	the_dir->field_width[0].len = len;
 	len += (flen = sprintf(the_dir->buf + len, "/%s", dirent->d_name));
-	the_dir->field_width[3] = flen-1;
+	the_dir->field_width[1].len = flen-1;
 
 	ftstr = ftype(dirent, iobuf->name);
 	len += (flen = sprintf(the_dir->buf + len, "/%s", ftstr));
-	the_dir->field_width[5] = flen-1;
+	the_dir->field_width[2].len = flen-1;
 
 	*out = the_dir->buf;
 
@@ -284,10 +285,10 @@ dir_take_control_of(awk_input_buf_t *iobuf)
 	emalloc(the_dir, open_directory_t *, sizeof(open_directory_t), "dir_take_control_of");
 	the_dir->dp = dp;
 	/* pre-populate the field_width array with constant values: */
-	the_dir->field_width[0] = 0;	/* no leading space */
-	the_dir->field_width[2] = 1;	/* single slash sign separator*/
-	the_dir->field_width[4] = 1;	/* single slash sign separator*/
-	the_dir->field_width[6] = -1;	/* terminate it after 3 fields */
+	the_dir->field_width[0].skip = 0;	/* no leading space */
+	the_dir->field_width[1].skip = 1;	/* single '/' separator */
+	the_dir->field_width[2].skip = 1;	/* single '/' separator */
+	the_dir->field_width[3].skip = -1;	/* terminate after 3 fields */
 	size = sizeof(struct dirent) + 21 /* max digits in inode */ + 2 /* slashes */;
 	emalloc(the_dir->buf, char *, size, "dir_take_control_of");
 
diff --git a/field.c b/field.c
index 5bdc05f7..5ef4d74b 100644
--- a/field.c
+++ b/field.c
@@ -59,7 +59,7 @@ static long fw_parse_field(long, char **, int, NODE *,
 			     Regexp *, Setfunc, NODE *, NODE *, bool);
 static long api_parse_field(long, char **, int, NODE *,
 			     Regexp *, Setfunc, NODE *, NODE *, bool);
-static const int *api_fw = NULL;
+static const awk_input_field_info_t *api_fw = NULL;
 static long fpat_parse_field(long, char **, int, NODE *,
 			     Regexp *, Setfunc, NODE *, NODE *, bool);
 static void set_element(long num, char * str, long len, NODE *arr);
@@ -262,7 +262,7 @@ rebuild_record()
  * but better correct than fast.
  */
 void
-set_record(const char *buf, int cnt, const int *fw)
+set_record(const char *buf, int cnt, const awk_input_field_info_t *fw)
 {
 	NODE *n;
 	static char *databuf;
@@ -802,24 +802,25 @@ api_parse_field(long up_to,	/* parse only up to this field number */
 	long nf = parse_high_water;
 	char *end = scan + len;
 	int skiplen;
+	size_t flen;
 
 	if (up_to == UNLIMITED)
 		nf = 0;
 	if (len == 0)
 		return nf;
 	while (nf < up_to) {
-		if (((skiplen = api_fw[2*nf]) < 0) ||
-		    ((len = api_fw[2*nf+1]) < 0)) {
+		if ((skiplen = api_fw[nf].skip) < 0) {
 		    	*buf = end;
 			return nf;
 		}
 		if (skiplen > end - scan)
 			skiplen = end - scan;
 		scan += skiplen;
-		if (len > end - scan)
-			len = end - scan;
-		(*set)(++nf, scan, (long) len, n);
-		scan += len;
+		flen = api_fw[nf].len;
+		if (flen > end - scan)
+			flen = end - scan;
+		(*set)(++nf, scan, (long) flen, n);
+		scan += flen;
 	}
 	*buf = scan;
 	return nf;
diff --git a/gawkapi.h b/gawkapi.h
index 6fa022ca..e744a0fc 100644
--- a/gawkapi.h
+++ b/gawkapi.h
@@ -117,6 +117,19 @@ typedef enum awk_bool {
 	awk_true
 } awk_bool_t;	/* we don't use <stdbool.h> on purpose */
 
+/*
+ * If the input parser would like to specify the field positions in the input
+ * record, it may populate an array of awk_input_field_info_t structures
+ * to indicate the location of each field. The 0th array element contains
+ * the information about field $1, and the NFth element should set skip
+ * to a negative value. For both skip and len, the value should be in
+ * bytes, not (potentially multi-byte) characters.
+ */
+typedef struct {
+	int skip;	/* # of bytes to skip before field starts */
+	size_t len;	/* # of bytes in field */
+} awk_input_field_info_t;
+
 /* The information about input files that input parsers need to know: */
 typedef struct awk_input {
 	const char *name;	/* filename */
@@ -149,12 +162,10 @@ typedef struct awk_input {
 	 *
 	 * If field_width is non-NULL, then its value will be initialized
 	 * to NULL, and the function may set it to point to an array of
-	 * integers supplying field width information to override the default
+	 * structures supplying field width information to override the default
 	 * gawk field parsing mechanism.  The field_width array should have
-	 * at least 2*NF+1 elements, and the value of field_width[2*NF]
-	 * must be negative. The first entry field_width[0] should contain
-	 * the number of bytes to skip before $1; field_width[1] contains
-	 * the number of bytes in $1. Note that these values are specified
+	 * at least NF+1 elements, and the value of field_width[NF].skip
+	 * must be negative. Note that these values are specified
 	 * in bytes, not (potentially multi-byte) characters! And note that this
 	 * array will not be copied by gawk; it must persist at least until the
 	 * next call to get_record or close_func. Note that field_width will
@@ -163,7 +174,7 @@ typedef struct awk_input {
 	 */
 	int (*get_record)(char **out, struct awk_input *iobuf, int *errcode,
 			char **rt_start, size_t *rt_len,
-			const int **field_width);
+			const awk_input_field_info_t **field_width);
 
 	/*
 	 * No argument prototype on read_func to allow for older systems
diff --git a/io.c b/io.c
index 8aa9da82..b049851f 100644
--- a/io.c
+++ b/io.c
@@ -287,7 +287,7 @@ static RECVALUE rsrescan(IOBUF *iop, struct recmatch *recm, SCANSTATE *state);
 
 static RECVALUE (*matchrec)(IOBUF *iop, struct recmatch *recm, SCANSTATE *state) = rs1scan;
 
-static int get_a_record(char **out, IOBUF *iop, int *errcode, const int **field_width);
+static int get_a_record(char **out, IOBUF *iop, int *errcode, const awk_input_field_info_t **field_width);
 
 static void free_rp(struct redirect *rp);
 
@@ -590,7 +590,7 @@ inrec(IOBUF *iop, int *errcode)
 	char *begin;
 	int cnt;
 	bool retval = true;
-	const int *field_width = NULL;
+	const awk_input_field_info_t *field_width = NULL;
 
 	if (at_eof(iop) && no_data_left(iop))
 		cnt = EOF;
@@ -2619,7 +2619,7 @@ do_getline_redir(int into_variable, enum redirval redirtype)
 	NODE *redir_exp = NULL;
 	NODE **lhs = NULL;
 	int redir_error = 0;
-	const int *field_width = NULL;
+	const awk_input_field_info_t *field_width = NULL;
 
 	if (into_variable)
 		lhs = POP_ADDRESS();
@@ -2688,7 +2688,7 @@ do_getline(int into_variable, IOBUF *iop)
 	int cnt = EOF;
 	char *s = NULL;
 	int errcode;
-	const int *field_width = NULL;
+	const awk_input_field_info_t *field_width = NULL;
 
 	if (iop == NULL) {	/* end of input */
 		if (into_variable)
@@ -3657,7 +3657,8 @@ static int
 get_a_record(char **out,        /* pointer to pointer to data */
         IOBUF *iop,             /* input IOP */
         int *errcode,           /* pointer to error variable */
-        const int **field_width)/* pointer to pointer to field_width array */
+        const awk_input_field_info_t **field_width)
+				/* pointer to pointer to field_width array */
 {
 	struct recmatch recm;
 	SCANSTATE state;