this repo has no description
more
+167
-57
+20
yaml/ocaml-yamle/bin/dune
+20
yaml/ocaml-yamle/bin/dune
···
6
6
(executable
7
7
(name test_emit)
8
8
(libraries yamle))
9
+
10
+
(executable
11
+
(name test_anchor)
12
+
(libraries yamle))
13
+
14
+
(executable
15
+
(name test_anchor_comprehensive)
16
+
(libraries yamle))
17
+
18
+
(executable
19
+
(name test_anchor_boundaries)
20
+
(libraries yamle))
21
+
22
+
(executable
23
+
(name test_alias)
24
+
(libraries yamle))
25
+
26
+
(executable
27
+
(name test_anchor_fix_final)
28
+
(libraries yamle))
+8
-4
yaml/ocaml-yamle/lib/parser.ml
+8
-4
yaml/ocaml-yamle/lib/parser.ml
···
101
101
102
102
(** Resolve a tag *)
103
103
let resolve_tag t ~handle ~suffix =
104
-
match List.assoc_opt handle t.tag_directives with
105
-
| Some prefix -> prefix ^ suffix
106
-
| None when handle = "!" -> "!" ^ suffix
107
-
| None -> Error.raise (Invalid_tag (handle ^ suffix))
104
+
if handle = "" then
105
+
(* Verbatim tag - suffix is already the full URI *)
106
+
suffix
107
+
else
108
+
match List.assoc_opt handle t.tag_directives with
109
+
| Some prefix -> prefix ^ suffix
110
+
| None when handle = "!" -> "!" ^ suffix
111
+
| None -> Error.raise (Invalid_tag (handle ^ suffix))
108
112
109
113
(** Process directives at document start *)
110
114
let process_directives t =
+123
-37
yaml/ocaml-yamle/lib/scanner.ml
+123
-37
yaml/ocaml-yamle/lib/scanner.ml
···
154
154
let scan_anchor_alias t =
155
155
let start = Input.mark t.input in
156
156
let buf = Buffer.create 16 in
157
+
(* Per YAML 1.2 spec: anchor names can contain any character that is NOT:
158
+
- Whitespace (space, tab, line breaks)
159
+
- Flow indicators: []{}
160
+
- Comma (,)
161
+
This matches the saphyr implementation: is_yaml_non_space && !is_flow *)
157
162
while
158
163
match Input.peek t.input with
159
-
| Some c when Input.is_alnum c || c = '_' || c = '-' ->
164
+
| Some c when not (Input.is_whitespace c) &&
165
+
not (Input.is_flow_indicator c) &&
166
+
c <> '\x00' ->
160
167
Buffer.add_char buf c;
161
168
ignore (Input.next t.input);
162
169
true
···
215
222
let handle, suffix =
216
223
match Input.peek t.input with
217
224
| Some '<' ->
218
-
(* Verbatim tag: !<...> *)
225
+
(* Verbatim tag: !<...> - handle is empty, suffix is full URI *)
219
226
ignore (Input.next t.input);
220
227
let buf = Buffer.create 32 in
221
228
while
···
228
235
| None -> Error.raise_at (Input.mark t.input) (Invalid_tag "unclosed verbatim tag")
229
236
do () done;
230
237
ignore (Input.next t.input); (* consume > *)
231
-
("!", Buffer.contents buf)
238
+
("", Buffer.contents buf)
232
239
| Some c when Input.is_whitespace c || Input.is_flow_indicator c ->
233
240
(* Non-specific tag: ! *)
234
241
("!", "")
···
238
245
let suffix = scan_tag_suffix t in
239
246
("!!", suffix)
240
247
| _ ->
241
-
(* Primary handle or just suffix *)
242
-
let first_part = scan_tag_suffix t in
243
-
if String.length first_part > 0 && first_part.[String.length first_part - 1] = '!' then
244
-
let suffix = scan_tag_suffix t in
245
-
(first_part, suffix)
246
-
else
247
-
("!", first_part)
248
+
(* Primary handle or just suffix: !foo or !e!foo *)
249
+
(* Read alphanumeric characters *)
250
+
let buf = Buffer.create 16 in
251
+
while
252
+
match Input.peek t.input with
253
+
| Some c when Input.is_alnum c || c = '-' ->
254
+
Buffer.add_char buf c;
255
+
ignore (Input.next t.input);
256
+
true
257
+
| _ -> false
258
+
do () done;
259
+
(* Check if next character is ! - if so, this is a named handle *)
260
+
(match Input.peek t.input with
261
+
| Some '!' ->
262
+
(* Named handle like !e! *)
263
+
ignore (Input.next t.input);
264
+
let handle_name = Buffer.contents buf in
265
+
let suffix = scan_tag_suffix t in
266
+
("!" ^ handle_name ^ "!", suffix)
267
+
| _ ->
268
+
(* Just ! followed by suffix *)
269
+
("!", Buffer.contents buf ^ scan_tag_suffix t))
248
270
in
249
271
let span = Span.make ~start ~stop:(Input.mark t.input) in
250
272
(handle, suffix, span)
···
471
493
scan_line ();
472
494
(* Check for line continuation *)
473
495
if not in_flow && Input.next_is_break t.input then begin
474
-
(* Save whitespace *)
475
-
Buffer.clear spaces;
476
-
Buffer.add_char spaces '\n';
496
+
(* Save the line break *)
497
+
if !leading_blanks then begin
498
+
(* We already had a break - this is an additional break (empty line) *)
499
+
Buffer.add_char spaces '\n'
500
+
end else begin
501
+
(* First line break *)
502
+
Buffer.clear spaces;
503
+
Buffer.add_char spaces '\n';
504
+
leading_blanks := true
505
+
end;
477
506
Input.consume_break t.input;
478
507
(* Line break in block context allows simple key *)
479
508
t.allow_simple_key <- true;
480
-
(* Skip leading blanks *)
509
+
(* Skip leading blanks on the next line *)
481
510
while Input.next_is_blank t.input do
482
-
Buffer.add_char spaces (Option.get (Input.next t.input))
511
+
ignore (Input.next t.input)
483
512
done;
484
513
let col = (Input.position t.input).column in
485
514
(* Check indentation - stop if we're at or before the containing block's indent *)
···
487
516
() (* Stop - dedented or at parent level *)
488
517
else if Input.at_document_boundary t.input then
489
518
() (* Stop - document boundary *)
490
-
else begin
491
-
leading_blanks := true;
519
+
else
492
520
scan_lines ()
493
-
end
494
521
end
495
522
in
496
523
···
562
589
let base_indent = current_indent t in
563
590
let content_indent = ref (
564
591
match !explicit_indent with
565
-
| Some n -> base_indent + n
592
+
| Some n ->
593
+
(* base_indent is a column (1-indexed), convert to indent level (0-indexed) *)
594
+
let base_level = max 0 (base_indent - 1) in
595
+
base_level + n
566
596
| None -> 0 (* Will be determined by first non-empty line *)
567
597
) in
568
598
569
599
let buf = Buffer.create 256 in
570
600
let trailing_breaks = Buffer.create 16 in
571
601
572
-
(* Read content *)
573
-
let rec read_lines () =
574
-
(* Skip empty lines, collecting breaks *)
575
-
while Input.next_is_break t.input ||
576
-
(Input.next_is_blank t.input &&
577
-
match Input.peek_nth t.input 1 with
578
-
| Some c when Input.is_break c -> true
579
-
| None -> true
580
-
| _ -> false)
581
-
do
582
-
if Input.next_is_blank t.input then begin
583
-
while Input.next_is_blank t.input do
584
-
ignore (Input.next t.input)
585
-
done
586
-
end;
602
+
(* Skip to content indentation, skipping empty lines *)
603
+
let rec skip_to_content_indent () =
604
+
if !content_indent > 0 then begin
605
+
(* Explicit indent - skip exactly content_indent spaces *)
606
+
let spaces_skipped = ref 0 in
607
+
while !spaces_skipped < !content_indent && Input.next_is (( = ) ' ') t.input do
608
+
incr spaces_skipped;
609
+
ignore (Input.next t.input)
610
+
done;
611
+
612
+
(* Check if this line is empty (only spaces/tabs until break/eof) *)
587
613
if Input.next_is_break t.input then begin
614
+
(* Empty line - record the break and continue *)
588
615
Buffer.add_char trailing_breaks '\n';
589
-
Input.consume_break t.input
616
+
Input.consume_break t.input;
617
+
skip_to_content_indent ()
618
+
end else if Input.next_is_blank t.input then begin
619
+
(* Line has spaces beyond content_indent - check if rest is only blanks *)
620
+
let idx = ref 0 in
621
+
let is_empty = ref false in
622
+
while not !is_empty do
623
+
match Input.peek_nth t.input !idx with
624
+
| Some c when Input.is_blank c -> incr idx
625
+
| Some c when Input.is_break c -> is_empty := true
626
+
| _ -> is_empty := true (* Not a break, so has content *)
627
+
done;
628
+
(* Check if we found a break (empty line) or content *)
629
+
(match Input.peek_nth t.input (!idx) with
630
+
| None | Some '\n' | Some '\r' ->
631
+
(* Empty line - consume all blanks and break *)
632
+
while Input.next_is_blank t.input do
633
+
ignore (Input.next t.input)
634
+
done;
635
+
Buffer.add_char trailing_breaks '\n';
636
+
Input.consume_break t.input;
637
+
skip_to_content_indent ()
638
+
| _ ->
639
+
(* Has content *)
640
+
())
590
641
end
591
-
done;
642
+
end else begin
643
+
(* Implicit indent - skip empty lines without consuming spaces *)
644
+
if Input.next_is_break t.input then begin
645
+
Buffer.add_char trailing_breaks '\n';
646
+
Input.consume_break t.input;
647
+
skip_to_content_indent ()
648
+
end else if Input.next_is_blank t.input then begin
649
+
(* Check if line is empty *)
650
+
let idx = ref 0 in
651
+
while match Input.peek_nth t.input !idx with
652
+
| Some c when Input.is_blank c -> incr idx; true
653
+
| _ -> false
654
+
do () done;
655
+
match Input.peek_nth t.input (!idx) with
656
+
| None | Some '\n' | Some '\r' ->
657
+
(* Empty line *)
658
+
while Input.next_is_blank t.input do
659
+
ignore (Input.next t.input)
660
+
done;
661
+
Buffer.add_char trailing_breaks '\n';
662
+
Input.consume_break t.input;
663
+
skip_to_content_indent ()
664
+
| _ ->
665
+
(* Has content - don't consume anything *)
666
+
()
667
+
end
668
+
end
669
+
in
670
+
671
+
(* Read content *)
672
+
let rec read_lines () =
673
+
skip_to_content_indent ();
592
674
593
675
(* Check if we're at content *)
594
676
if Input.is_eof t.input then ()
···
601
683
ignore (Input.next t.input)
602
684
done;
603
685
604
-
(* Determine content indent from first content line *)
686
+
(* For explicit indent, line_indent is extra beyond content_indent *)
687
+
if !content_indent > 0 then
688
+
line_indent := !content_indent + !line_indent;
689
+
690
+
(* Determine content indent from first content line (implicit case) *)
605
691
if !content_indent = 0 then begin
606
692
if !line_indent <= base_indent then begin
607
693
(* No content - restore position conceptually *)
+1
-1
yaml/ocaml-yamle/tests/cram/anchors.t
+1
-1
yaml/ocaml-yamle/tests/cram/anchors.t
+12
-12
yaml/ocaml-yamle/tests/cram/comments.t
+12
-12
yaml/ocaml-yamle/tests/cram/comments.t
···
1
1
Test comments.yml file with various comment styles
2
2
3
3
$ yamlcat ../yaml/comments.yml
4
-
name: John Doe
5
-
age: 30
4
+
name: John Doe
5
+
age: 30
6
6
address:
7
-
street: 123 Main St
7
+
street: 123 Main St
8
8
city: Springfield
9
9
zip: 12345
10
10
items:
11
-
- apple
11
+
- apple
12
12
- banana
13
13
- cherry
14
-
- date
14
+
- date
15
15
flow_seq:
16
16
- 1
17
17
- 2
···
22
22
nested:
23
23
level1:
24
24
level2:
25
-
value: deeply nested
25
+
value: deeply nested
26
26
multi_comment_key: value
27
27
special: 'value with # hash inside quotes'
28
28
empty_value: null
···
31
31
Test comments.yml roundtrip with JSON to verify parsed values
32
32
33
33
$ yamlcat --json ../yaml/comments.yml
34
-
{"name": "John Doe ", "age": "30 ", "address": {"street": "123 Main St ", "city": "Springfield", "zip": 12345}, "items": ["apple ", "banana", "cherry", "date "], "flow_seq": [1, 2, 3], "flow_map": {"key1": "value1", "key2": "value2"}, "nested": {"level1": {"level2": {"value": "deeply nested "}}}, "multi_comment_key": "value", "special": "value with # hash inside quotes", "empty_value": null, "final_key": "final_value"}
34
+
{"name": "John Doe", "age": 30, "address": {"street": "123 Main St", "city": "Springfield", "zip": 12345}, "items": ["apple", "banana", "cherry", "date"], "flow_seq": [1, 2, 3], "flow_map": {"key1": "value1", "key2": "value2"}, "nested": {"level1": {"level2": {"value": "deeply nested"}}}, "multi_comment_key": "value", "special": "value with # hash inside quotes", "empty_value": null, "final_key": "final_value"}
35
35
36
36
Test full line comments are ignored
37
37
···
45
45
46
46
$ echo 'name: Bob # This is an end of line comment
47
47
> age: 35 # Another end of line comment' | yamlcat --json
48
-
{"name": "Bob ", "age": "35 "}
48
+
{"name": "Bob", "age": 35}
49
49
50
50
Test comments after sequence items
51
51
···
53
53
> - apple # First fruit
54
54
> - banana # Second fruit
55
55
> - cherry # Third fruit' | yamlcat --json
56
-
{"fruits": ["apple ", "banana ", "cherry "]}
56
+
{"fruits": ["apple", "banana", "cherry"]}
57
57
58
58
Test comments between sequence items
59
59
···
99
99
> inner:
100
100
> # Comment in deeper level
101
101
> key: value # End of line comment' | yamlcat --json
102
-
{"outer": {"inner": {"key": "value "}}}
102
+
{"outer": {"inner": {"key": "value"}}}
103
103
104
104
Test comments with special characters
105
105
···
158
158
> # Comment between sequence items
159
159
> - name: item2
160
160
> value: 20 # Another comment' | yamlcat --json
161
-
{"items": [{"name": "item1 ", "value": 10}, {"name": "item2", "value": "20 "}]}
161
+
{"items": [{"name": "item1", "value": 10}, {"name": "item2", "value": 20}]}
162
162
163
163
Test comment only lines between complex structures
164
164
···
188
188
Test whitespace preservation around comments
189
189
190
190
$ echo 'key1: value1 # Comment with spaces' | yamlcat --json
191
-
{"key1": "value1 "}
191
+
{"key1": "value1"}
192
192
193
193
Test comment after colon but before value
194
194
+1
-1
yaml/ocaml-yamle/tests/cram/multidoc.t
+1
-1
yaml/ocaml-yamle/tests/cram/multidoc.t
+2
-2
yaml/ocaml-yamle/tests/cram/yamlcat.t
+2
-2
yaml/ocaml-yamle/tests/cram/yamlcat.t
···
56
56
Test seq.yml file (multiline plain scalar)
57
57
58
58
$ yamlcat ../yaml/seq.yml
59
-
- hello - whats - up
59
+
- hello - whats - up
60
60
- foo
61
61
- bar
62
62
63
63
Test seq.yml roundtrip preserves data
64
64
65
65
$ yamlcat --json ../yaml/seq.yml
66
-
["hello - whats - up", "foo", "bar"]
66
+
["hello - whats - up", "foo", "bar"]
67
67
68
68
Test cohttp.yml
69
69