From: Ben Pfaff <blp@cs.stanford.edu>
Date: Sun, 4 Jul 2021 05:00:47 +0000 (-0700)
Subject: segment: Distinguish snippets from full files.
X-Git-Url: https://pintos-os.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=dc25c013a7ce573b87fd27cbefddc732733c837f;p=pspp

segment: Distinguish snippets from full files.

The comment on segmenter_init() explains what this means:

If IS_SNIPPET is false, then the segmenter will parse as if it's being
given a whole file.  This means, for example, that it will interpret -
or + at the beginning of the syntax as a separator between commands
(since - or + at the beginning of a line has this meaning).

If IS_SNIPPET is true, then the segmenter will parse as if it's being
given an isolated piece of syntax.  This means that, for example, that
it will interpret - or + at the beginning of the syntax as an operator
token or (if followed by a digit) as part of a number.
---

diff --git a/src/language/control/repeat.c b/src/language/control/repeat.c
index 118e8d3ccd..86dd36f7f0 100644
--- a/src/language/control/repeat.c
+++ b/src/language/control/repeat.c
@@ -201,10 +201,7 @@ do_parse_commands (struct substring s, enum segmenter_mode mode,
                    struct hmap *dummies,
                    struct string *outputs, size_t n_outputs)
 {
-  struct segmenter segmenter;
-
-  segmenter_init (&segmenter, mode);
-
+  struct segmenter segmenter = segmenter_init (mode, false);
   while (!ss_is_empty (s))
     {
       enum segment_type type;
diff --git a/src/language/lexer/lexer.c b/src/language/lexer/lexer.c
index df2806eedf..718458f8da 100644
--- a/src/language/lexer/lexer.c
+++ b/src/language/lexer/lexer.c
@@ -988,7 +988,7 @@ lex_match_phrase (struct lexer *lexer, const char *s)
   int i;
 
   i = 0;
-  string_lexer_init (&slex, s, strlen (s), SEG_MODE_INTERACTIVE);
+  string_lexer_init (&slex, s, strlen (s), SEG_MODE_INTERACTIVE, true);
   while (string_lexer_next (&slex, &token))
     if (token.type != SCAN_SKIP)
       {
@@ -1238,7 +1238,8 @@ lex_interactive_reset (struct lexer *lexer)
       src->journal_pos = src->seg_pos = src->line_pos = 0;
       src->n_newlines = 0;
       src->suppress_next_newline = false;
-      segmenter_init (&src->segmenter, segmenter_get_mode (&src->segmenter));
+      src->segmenter = segmenter_init (segmenter_get_mode (&src->segmenter),
+                                       false);
       while (!deque_is_empty (&src->deque))
         lex_source_pop__ (src);
       lex_source_push_endcmd__ (src);
@@ -1658,12 +1659,12 @@ lex_source_push_endcmd__ (struct lex_source *src)
 static struct lex_source *
 lex_source_create (struct lex_reader *reader)
 {
-  struct lex_source *src;
-
-  src = xzalloc (sizeof *src);
-  src->reader = reader;
-  segmenter_init (&src->segmenter, reader->syntax);
-  src->tokens = deque_init (&src->deque, 4, sizeof *src->tokens);
+  struct lex_source *src = xmalloc (sizeof *src);
+  *src = (struct lex_source) {
+    .reader = reader,
+    .segmenter = segmenter_init (reader->syntax, false),
+    .tokens = deque_init (&src->deque, 4, sizeof *src->tokens),
+  };
 
   lex_source_push_endcmd__ (src);
 
diff --git a/src/language/lexer/scan.c b/src/language/lexer/scan.c
index 0c92210540..0acff46cdd 100644
--- a/src/language/lexer/scan.c
+++ b/src/language/lexer/scan.c
@@ -651,12 +651,14 @@ scanner_push (struct scanner *scanner, enum segment_type type,
    INPUT must not be modified or freed while SLEX is still in use. */
 void
 string_lexer_init (struct string_lexer *slex, const char *input, size_t length,
-                   enum segmenter_mode mode)
+                   enum segmenter_mode mode, bool is_snippet)
 {
-  slex->input = input;
-  slex->length = length;
-  slex->offset = 0;
-  segmenter_init (&slex->segmenter, mode);
+  *slex = (struct string_lexer) {
+    .input = input,
+    .length = length,
+    .offset = 0,
+    .segmenter = segmenter_init (mode, is_snippet),
+  };
 }
 
 /*  */
diff --git a/src/language/lexer/scan.h b/src/language/lexer/scan.h
index 61bfc5b553..0dde273804 100644
--- a/src/language/lexer/scan.h
+++ b/src/language/lexer/scan.h
@@ -102,7 +102,7 @@ struct string_lexer
   };
 
 void string_lexer_init (struct string_lexer *, const char *input,
-                        size_t length, enum segmenter_mode);
+                        size_t length, enum segmenter_mode, bool is_snippet);
 bool string_lexer_next (struct string_lexer *, struct token *);
 
 #endif /* scan.h */
diff --git a/src/language/lexer/segment.c b/src/language/lexer/segment.c
index 35240b4c64..2689811593 100644
--- a/src/language/lexer/segment.c
+++ b/src/language/lexer/segment.c
@@ -1786,17 +1786,28 @@ segment_type_to_string (enum segment_type type)
     }
 }
 
-/* Initializes S as a segmenter with the given syntax MODE.
+/* Returns a segmenter with the given syntax MODE.
+
+   If IS_SNIPPET is false, then the segmenter will parse as if it's being given
+   a whole file.  This means, for example, that it will interpret - or + at the
+   beginning of the syntax as a separator between commands (since - or + at the
+   beginning of a line has this meaning).
+
+   If IS_SNIPPET is true, then the segmenter will parse as if it's being given
+   an isolated piece of syntax.  This means that, for example, that it will
+   interpret - or + at the beginning of the syntax as an operator token or (if
+   followed by a digit) as part of a number.
 
    A segmenter does not contain any external references, so nothing needs to be
    done to destroy one.  For the same reason, segmenters may be copied with
    plain struct assignment (or memcpy). */
-void
-segmenter_init (struct segmenter *s, enum segmenter_mode mode)
+struct segmenter
+segmenter_init (enum segmenter_mode mode, bool is_snippet)
 {
-  s->state = S_SHBANG;
-  s->substate = 0;
-  s->mode = mode;
+  return (struct segmenter) {
+    .state = is_snippet ? S_GENERAL : S_SHBANG,
+    .mode = mode,
+  };
 }
 
 /* Returns the mode passed to segmenter_init() for S. */
diff --git a/src/language/lexer/segment.h b/src/language/lexer/segment.h
index 02a269bdd2..5d550f531f 100644
--- a/src/language/lexer/segment.h
+++ b/src/language/lexer/segment.h
@@ -117,7 +117,7 @@ struct segmenter
     unsigned char mode;
   };
 
-void segmenter_init (struct segmenter *, enum segmenter_mode);
+struct segmenter segmenter_init (enum segmenter_mode, bool is_snippet);
 
 enum segmenter_mode segmenter_get_mode (const struct segmenter *);
 
diff --git a/tests/language/lexer/scan-test.c b/tests/language/lexer/scan-test.c
index 1eb04338e3..2a77e127ac 100644
--- a/tests/language/lexer/scan-test.c
+++ b/tests/language/lexer/scan-test.c
@@ -73,7 +73,7 @@ main (int argc, char *argv[])
         length--;
     }
 
-  string_lexer_init (&slex, input, length, mode);
+  string_lexer_init (&slex, input, length, mode, false);
   do
     {
       struct token token;
diff --git a/tests/language/lexer/segment-test.c b/tests/language/lexer/segment-test.c
index a3b67b89b2..acb444f200 100644
--- a/tests/language/lexer/segment-test.c
+++ b/tests/language/lexer/segment-test.c
@@ -108,8 +108,7 @@ main (int argc, char *argv[])
 static void
 check_segmentation (const char *input, size_t length, bool print_segments)
 {
-  struct segmenter s;
-  segmenter_init (&s, mode);
+  struct segmenter s = segmenter_init (mode, false);
 
   size_t line_number = 1;
   size_t line_offset = 0;