Skip to content
This repository was archived by the owner on Jun 15, 2023. It is now read-only.

Commit a612f22

Browse files
IwanKaramazowIwanKaramazow
authored
Expose API to seed the initial line number of the parser. (#65)
* Expose API to seed the initial line number of the parser. * Add extra documentation for line param in parser interface. * Add extra test for Napkin_parser.make with default line number
1 parent 341c6b9 commit a612f22

File tree

7 files changed

+207
-180
lines changed

7 files changed

+207
-180
lines changed

‎.depend

Lines changed: 3 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -75,5 +75,6 @@ src/napkin_scanner.cmx : src/napkin_token.cmx src/napkin_diagnostics.cmx \
7575
src/napkin_scanner.cmi
7676
src/napkin_scanner.cmi : src/napkin_token.cmx src/napkin_diagnostics.cmi
7777
src/napkin_token.cmx : src/napkin_comment.cmx src/napkin_character_codes.cmx
78-
tests/napkin_test.cmx : src/napkin_outcome_printer.cmx \
79-
src/napkin_multi_printer.cmx src/napkin_io.cmx src/napkin_driver.cmx
78+
tests/napkin_test.cmx : src/napkin_token.cmx src/napkin_parser.cmx \
79+
src/napkin_outcome_printer.cmx src/napkin_multi_printer.cmx \
80+
src/napkin_io.cmx src/napkin_driver.cmx

‎src/napkin_parser.ml

Lines changed: 145 additions & 147 deletions
Original file line numberDiff line numberDiff line change
@@ -8,152 +8,150 @@ module Comment = Napkin_comment
88

99
type mode = ParseForTypeChecker | Default
1010

11-
type regionStatus = Report | Silent
12-
13-
type t = {
14-
mode: mode;
15-
mutable scanner: Scanner.t;
16-
mutable token: Token.t;
17-
mutable startPos: Lexing.position;
18-
mutable endPos: Lexing.position;
19-
mutable prevEndPos: Lexing.position;
20-
mutable breadcrumbs: (Grammar.t * Lexing.position) list;
21-
mutable errors: Reporting.parseError list;
22-
mutable diagnostics: Diagnostics.t list;
23-
mutable comments: Comment.t list;
24-
mutable regions: regionStatus ref list;
25-
}
26-
27-
let err ?startPos ?endPos p error =
28-
let d = Diagnostics.make
29-
~filename:p.scanner.filename
30-
~startPos:(match startPos with | Some pos -> pos | None -> p.startPos)
31-
~endPos:(match endPos with | Some pos -> pos | None -> p.endPos)
11+
type regionStatus = Report | Silent
12+
13+
type t = {
14+
mode: mode;
15+
mutable scanner: Scanner.t;
16+
mutable token: Token.t;
17+
mutable startPos: Lexing.position;
18+
mutable endPos: Lexing.position;
19+
mutable prevEndPos: Lexing.position;
20+
mutable breadcrumbs: (Grammar.t * Lexing.position) list;
21+
mutable errors: Reporting.parseError list;
22+
mutable diagnostics: Diagnostics.t list;
23+
mutable comments: Comment.t list;
24+
mutable regions: regionStatus ref list;
25+
}
26+
27+
let err ?startPos ?endPos p error =
28+
let d = Diagnostics.make
29+
~filename:p.scanner.filename
30+
~startPos:(match startPos with | Some pos -> pos | None -> p.startPos)
31+
~endPos:(match endPos with | Some pos -> pos | None -> p.endPos)
32+
error
33+
in
34+
try
35+
if (!(List.hd p.regions) = Report) then (
36+
p.diagnostics <- d::p.diagnostics;
37+
List.hd p.regions := Silent
38+
)
39+
with Failure _ -> ()
40+
41+
let beginRegion p =
42+
p.regions <- ref Report :: p.regions
43+
let endRegion p =
44+
try p.regions <- List.tl p.regions with Failure _ -> ()
45+
46+
(* Advance to the next non-comment token and store any encountered comment
47+
* in the parser's state. Every comment contains the end position of its
48+
* previous token to facilite comment interleaving *)
49+
let rec next ?prevEndPos p =
50+
let prevEndPos = match prevEndPos with Some pos -> pos | None -> p.endPos in
51+
let (startPos, endPos, token) = Scanner.scan p.scanner in
52+
match token with
53+
| Comment c ->
54+
Comment.setPrevTokEndPos c p.endPos;
55+
p.comments <- c::p.comments;
56+
p.prevEndPos <- p.endPos;
57+
p.endPos <- endPos;
58+
next ~prevEndPos p
59+
| _ ->
60+
p.token <- token;
61+
(* p.prevEndPos <- prevEndPos; *)
62+
p.prevEndPos <- prevEndPos;
63+
p.startPos <- startPos;
64+
p.endPos <- endPos
65+
66+
let checkProgress ~prevEndPos ~result p =
67+
if p.endPos == prevEndPos
68+
then None
69+
else Some result
70+
71+
let make ?(mode=ParseForTypeChecker) ?line src filename =
72+
let scanner = Scanner.make ~filename ?line (Bytes.of_string src) in
73+
let parserState = {
74+
mode;
75+
scanner;
76+
token = Token.Eof;
77+
startPos = Lexing.dummy_pos;
78+
prevEndPos = Lexing.dummy_pos;
79+
endPos = Lexing.dummy_pos;
80+
breadcrumbs = [];
81+
errors = [];
82+
diagnostics = [];
83+
comments = [];
84+
regions = [ref Report];
85+
} in
86+
parserState.scanner.err <- (fun ~startPos ~endPos error ->
87+
let diagnostic = Diagnostics.make
88+
~filename
89+
~startPos
90+
~endPos
3291
error
3392
in
34-
try
35-
if (!(List.hd p.regions) = Report) then (
36-
p.diagnostics <- d::p.diagnostics;
37-
List.hd p.regions := Silent
38-
)
39-
with Failure _ -> ()
40-
41-
let beginRegion p =
42-
p.regions <- ref Report :: p.regions
43-
let endRegion p =
44-
try p.regions <- List.tl p.regions with Failure _ -> ()
45-
46-
(* Advance to the next non-comment token and store any encountered comment
47-
* in the parser's state. Every comment contains the end position of its
48-
* previous token to facilite comment interleaving *)
49-
let rec next ?prevEndPos p =
50-
let prevEndPos = match prevEndPos with Some pos -> pos | None -> p.endPos in
51-
let (startPos, endPos, token) = Scanner.scan p.scanner in
52-
match token with
53-
| Comment c ->
54-
Comment.setPrevTokEndPos c p.endPos;
55-
p.comments <- c::p.comments;
56-
p.prevEndPos <- p.endPos;
57-
p.endPos <- endPos;
58-
next ~prevEndPos p
59-
| _ ->
60-
p.token <- token;
61-
(* p.prevEndPos <- prevEndPos; *)
62-
p.prevEndPos <- prevEndPos;
63-
p.startPos <- startPos;
64-
p.endPos <- endPos
65-
66-
let checkProgress ~prevEndPos ~result p =
67-
if p.endPos == prevEndPos
68-
then None
69-
else Some result
70-
71-
let make ?(mode=ParseForTypeChecker) src filename =
72-
let scanner = Scanner.make (Bytes.of_string src) filename in
73-
let parserState = {
74-
mode;
75-
scanner;
76-
token = Token.Eof;
77-
startPos = Lexing.dummy_pos;
78-
prevEndPos = Lexing.dummy_pos;
79-
endPos = Lexing.dummy_pos;
80-
breadcrumbs = [];
81-
errors = [];
82-
diagnostics = [];
83-
comments = [];
84-
regions = [ref Report];
85-
} in
86-
parserState.scanner.err <- (fun ~startPos ~endPos error ->
87-
let diagnostic = Diagnostics.make
88-
~filename
89-
~startPos
90-
~endPos
91-
error
92-
in
93-
parserState.diagnostics <- diagnostic::parserState.diagnostics
94-
);
95-
next parserState;
96-
parserState
97-
98-
let leaveBreadcrumb p circumstance =
99-
let crumb = (circumstance, p.startPos) in
100-
p.breadcrumbs <- crumb::p.breadcrumbs
101-
102-
let eatBreadcrumb p =
103-
match p.breadcrumbs with
104-
| [] -> ()
105-
| _::crumbs -> p.breadcrumbs <- crumbs
106-
107-
let optional p token =
108-
if p.token = token then
109-
let () = next p in true
110-
else
111-
false
112-
113-
let expect ?grammar token p =
114-
if p.token = token then
115-
next p
116-
else
117-
let error = Diagnostics.expected ?grammar p.prevEndPos token in
118-
err ~startPos:p.prevEndPos p error
119-
120-
(* Don't use immutable copies here, it trashes certain heuristics
121-
* in the ocaml compiler, resulting in massive slowdowns of the parser *)
122-
let lookahead p callback =
123-
let err = p.scanner.err in
124-
let ch = p.scanner.ch in
125-
let offset = p.scanner.offset in
126-
let rdOffset = p.scanner.rdOffset in
127-
let lineOffset = p.scanner.lineOffset in
128-
let lnum = p.scanner.lnum in
129-
let mode = p.scanner.mode in
130-
let token = p.token in
131-
let startPos = p.startPos in
132-
let endPos = p.endPos in
133-
let prevEndPos = p.prevEndPos in
134-
let breadcrumbs = p.breadcrumbs in
135-
let errors = p.errors in
136-
let diagnostics = p.diagnostics in
137-
let comments = p.comments in
138-
139-
let res = callback p in
140-
141-
p.scanner.err <- err;
142-
p.scanner.ch <- ch;
143-
p.scanner.offset <- offset;
144-
p.scanner.rdOffset <- rdOffset;
145-
p.scanner.lineOffset <- lineOffset;
146-
p.scanner.lnum <- lnum;
147-
p.scanner.mode <- mode;
148-
p.token <- token;
149-
p.startPos <- startPos;
150-
p.endPos <- endPos;
151-
p.prevEndPos <- prevEndPos;
152-
p.breadcrumbs <- breadcrumbs;
153-
p.errors <- errors;
154-
p.diagnostics <- diagnostics;
155-
p.comments <- comments;
156-
157-
res
158-
159-
93+
parserState.diagnostics <- diagnostic::parserState.diagnostics
94+
);
95+
next parserState;
96+
parserState
97+
98+
let leaveBreadcrumb p circumstance =
99+
let crumb = (circumstance, p.startPos) in
100+
p.breadcrumbs <- crumb::p.breadcrumbs
101+
102+
let eatBreadcrumb p =
103+
match p.breadcrumbs with
104+
| [] -> ()
105+
| _::crumbs -> p.breadcrumbs <- crumbs
106+
107+
let optional p token =
108+
if p.token = token then
109+
let () = next p in true
110+
else
111+
false
112+
113+
let expect ?grammar token p =
114+
if p.token = token then
115+
next p
116+
else
117+
let error = Diagnostics.expected ?grammar p.prevEndPos token in
118+
err ~startPos:p.prevEndPos p error
119+
120+
(* Don't use immutable copies here, it trashes certain heuristics
121+
* in the ocaml compiler, resulting in massive slowdowns of the parser *)
122+
let lookahead p callback =
123+
let err = p.scanner.err in
124+
let ch = p.scanner.ch in
125+
let offset = p.scanner.offset in
126+
let rdOffset = p.scanner.rdOffset in
127+
let lineOffset = p.scanner.lineOffset in
128+
let lnum = p.scanner.lnum in
129+
let mode = p.scanner.mode in
130+
let token = p.token in
131+
let startPos = p.startPos in
132+
let endPos = p.endPos in
133+
let prevEndPos = p.prevEndPos in
134+
let breadcrumbs = p.breadcrumbs in
135+
let errors = p.errors in
136+
let diagnostics = p.diagnostics in
137+
let comments = p.comments in
138+
139+
let res = callback p in
140+
141+
p.scanner.err <- err;
142+
p.scanner.ch <- ch;
143+
p.scanner.offset <- offset;
144+
p.scanner.rdOffset <- rdOffset;
145+
p.scanner.lineOffset <- lineOffset;
146+
p.scanner.lnum <- lnum;
147+
p.scanner.mode <- mode;
148+
p.token <- token;
149+
p.startPos <- startPos;
150+
p.endPos <- endPos;
151+
p.prevEndPos <- prevEndPos;
152+
p.breadcrumbs <- breadcrumbs;
153+
p.errors <- errors;
154+
p.diagnostics <- diagnostics;
155+
p.comments <- comments;
156+
157+
res

‎src/napkin_parser.mli

Lines changed: 3 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -22,7 +22,8 @@ type t = {
2222
mutable regions: regionStatus ref list;
2323
}
2424

25-
val make: ?mode: mode -> string -> string -> t
25+
(* `line` seeds the parser's state with an initial line number. *)
26+
val make: ?mode:mode -> ?line:int -> string -> string -> t
2627

2728
val expect: ?grammar:Grammar.t -> Token.t -> t -> unit
2829
val optional: t -> Token.t -> bool
@@ -43,4 +44,4 @@ val checkProgress:
4344
prevEndPos: Lexing.position ->
4445
result: 'a ->
4546
t ->
46-
'a option
47+
'a option

‎src/napkin_reason_binary_driver.ml

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -15,7 +15,7 @@ let extractConcreteSyntax filename =
1515
if String.length filename > 0 then IO.readFile ~filename
1616
else IO.readStdin ()
1717
in
18-
let scanner = Napkin_scanner.make (Bytes.of_string src) filename in
18+
let scanner = Napkin_scanner.make (Bytes.of_string src) ~filename in
1919

2020
let rec next prevEndPos scanner =
2121
let (startPos, endPos, token) = Napkin_scanner.scan scanner in

‎src/napkin_scanner.ml

Lines changed: 3 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -77,7 +77,7 @@ let peek scanner =
7777
else
7878
-1
7979

80-
let make b filename =
80+
let make ?(line=1) ~filename b =
8181
let scanner = {
8282
filename;
8383
src = b;
@@ -86,7 +86,7 @@ let make b filename =
8686
offset = 0;
8787
rdOffset = 0;
8888
lineOffset = 0;
89-
lnum = 1;
89+
lnum = line;
9090
mode = [];
9191
} in
9292
next scanner;
@@ -726,4 +726,4 @@ let isBinaryOp src startCnum endCnum =
726726
CharacterCodes.isLineBreak c ||
727727
c == CharacterCodes.eof
728728
in
729-
leftOk && rightOk
729+
leftOk && rightOk

0 commit comments

Comments
 (0)