fix(parser): emit lexer error on escaped surrogate to avoid panic (#658)

* fix(parser): emit lexer error on escaped surrogate to avoid panic Fixes #608 * Changelog
apollographql · Sep 28, 2023 · 7719507 · 7719507
1 parent a922484
commit 7719507
Show file tree

Hide file tree

Showing 8 changed files with 66 additions and 20 deletions.
diff --git a/crates/apollo-parser/CHANGELOG.md b/crates/apollo-parser/CHANGELOG.md
@@ -19,7 +19,7 @@ This project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.htm
 
 # [x.x.x] (unreleased) - 2023-mm-dd
 ## Fixes
-- **apply recursion limit where needed, reduce its default from 4096 to 500 [SimonSapin], [pull/662]**
+- **apply recursion limit where needed, reduce its default from 4096 to 500 - [SimonSapin], [pull/662]**
   The limit was only tracked for nested selection sets, but the parser turns out
   to use recursion in other cases too. [Issue 666] tracks reducing them.
   Stack overflow was observed with little more than 2000
@@ -28,9 +28,12 @@ This project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.htm
 - **fix various lexer bugs - [SimonSapin], [pull/646], [pull/652]**
   The lexer was too permissive in emitting tokens instead of errors
   in various cases around numbers, strings, and EOF.
+- **fix panic on surrogate code points in unicode escape sequences - [SimonSapin], [issue/608], [pull/658]**
 
+[issue/608]: https://github.com/apollographql/apollo-rs/issues/608
 [pull/646]: https://github.com/apollographql/apollo-rs/pull/646
 [pull/652]: https://github.com/apollographql/apollo-rs/pull/652
+[pull/658]: https://github.com/apollographql/apollo-rs/pull/658
 [pull/662]: https://github.com/apollographql/apollo-rs/pull/662
 [Issue 666]: https://github.com/apollographql/apollo-rs/issues/666
 

diff --git a/crates/apollo-parser/src/ast/node_ext.rs b/crates/apollo-parser/src/ast/node_ext.rs
@@ -167,6 +167,8 @@ fn unescape_string(input: &str) -> String {
                     break;
                 };
 
+                // TODO: https://github.com/apollographql/apollo-rs/issues/657 needs
+                // changes both here and in `lexer/mod.rs`
                 let mut unicode = || {
                     // 1. Let value be the 16-bit hexadecimal value represented
                     // by the sequence of hexadecimal digits within EscapedUnicode.

diff --git a/crates/apollo-parser/src/lexer/cursor.rs b/crates/apollo-parser/src/lexer/cursor.rs
@@ -6,8 +6,8 @@ use crate::Error;
 #[derive(Debug, Clone)]
 pub(crate) struct Cursor<'a> {
     index: usize,
-    offset: usize,
-    source: &'a str,
+    pub(super) offset: usize,
+    pub(super) source: &'a str,
     chars: CharIndices<'a>,
     pending: Option<char>,
     pub(crate) err: Option<Error>,

diff --git a/crates/apollo-parser/src/lexer/mod.rs b/crates/apollo-parser/src/lexer/mod.rs
@@ -326,7 +326,24 @@ impl<'a> Cursor<'a> {
                     _ => {
                         if remaining <= 1 {
                             state = State::StringLiteral;
-
+                            let hex_end = self.offset + 1;
+                            let hex_start = hex_end - 4;
+                            let hex = &self.source[hex_start..hex_end];
+                            // `is_ascii_hexdigit()` checks in previous iterations ensures
+                            // this `unwrap()` does not panic:
+                            let code_point = u32::from_str_radix(hex, 16).unwrap();
+                            if char::from_u32(code_point).is_none() {
+                                // TODO: https://github.com/apollographql/apollo-rs/issues/657 needs
+                                // changes both here and in `ast/node_ext.rs`
+                                let escape_sequence_start = hex_start - 2; // include "\u"
+                                let escape_sequence = &self.source[escape_sequence_start..hex_end];
+                                self.add_err(Error::new(
+                                    "surrogate code point is invalid in unicode escape sequence \
+                                     (paired surrogate not supported yet: \
+                                     https://github.com/apollographql/apollo-rs/issues/657)",
+                                    escape_sequence.to_owned(),
+                                ));
+                            }
                             continue;
                         }
 

diff --git a/crates/apollo-parser/test_data/lexer/err/0030_escaped_surrogate.graphql b/crates/apollo-parser/test_data/lexer/err/0030_escaped_surrogate.graphql
@@ -0,0 +1,12 @@
+# TODO: move these cases back to ok/0004_string_value.graphql when
+# https://github.com/apollographql/apollo-rs/issues/657 is implemented
+"string with unicode surrogate pair escape \uD83D\uDE00"
+"string with minimal surrogate pair escape \uD800\uDC00"
+"string with maximal surrogate pair escape \uDBFF\uDFFF"
+
+# TODO: emit two errors: https://github.com/apollographql/apollo-rs/issues/319
+"split pair \uD83D \uDE00"
+
+"Backwards pair \uDE00\uD83D"
+"Lone lead surrogate \uD83E"
+"Lone trail surrogate \uDD80"
diff --git a/crates/apollo-parser/test_data/lexer/err/0030_escaped_surrogate.txt b/crates/apollo-parser/test_data/lexer/err/0030_escaped_surrogate.txt
@@ -0,0 +1,21 @@
+COMMENT@0:66 "# TODO: move these cases back to ok/0004_string_value.graphql when"
+WHITESPACE@66:67 "\n"
+COMMENT@67:137 "# https://github.com/apollographql/apollo-rs/issues/657 is implemented"
+WHITESPACE@137:138 "\n"
+ERROR@138:194 "surrogate code point is invalid in unicode escape sequence (paired surrogate not supported yet: https://github.com/apollographql/apollo-rs/issues/657)" "string with unicode surrogate pair escape \uD83D\uDE00"
+WHITESPACE@194:195 "\n"
+ERROR@195:251 "surrogate code point is invalid in unicode escape sequence (paired surrogate not supported yet: https://github.com/apollographql/apollo-rs/issues/657)" "string with minimal surrogate pair escape \uD800\uDC00"
+WHITESPACE@251:252 "\n"
+ERROR@252:308 "surrogate code point is invalid in unicode escape sequence (paired surrogate not supported yet: https://github.com/apollographql/apollo-rs/issues/657)" "string with maximal surrogate pair escape \uDBFF\uDFFF"
+WHITESPACE@308:310 "\n\n"
+COMMENT@310:388 "# TODO: emit two errors: https://github.com/apollographql/apollo-rs/issues/319"
+WHITESPACE@388:389 "\n"
+ERROR@389:415 "surrogate code point is invalid in unicode escape sequence (paired surrogate not supported yet: https://github.com/apollographql/apollo-rs/issues/657)" "split pair \uD83D \uDE00"
+WHITESPACE@415:417 "\n\n"
+ERROR@417:446 "surrogate code point is invalid in unicode escape sequence (paired surrogate not supported yet: https://github.com/apollographql/apollo-rs/issues/657)" "Backwards pair \uDE00\uD83D"
+WHITESPACE@446:447 "\n"
+ERROR@447:475 "surrogate code point is invalid in unicode escape sequence (paired surrogate not supported yet: https://github.com/apollographql/apollo-rs/issues/657)" "Lone lead surrogate \uD83E"
+WHITESPACE@475:476 "\n"
+ERROR@476:505 "surrogate code point is invalid in unicode escape sequence (paired surrogate not supported yet: https://github.com/apollographql/apollo-rs/issues/657)" "Lone trail surrogate \uDD80"
+WHITESPACE@505:506 "\n"
+EOF@506:506
diff --git a/crates/apollo-parser/test_data/lexer/ok/0004_string_value.graphql b/crates/apollo-parser/test_data/lexer/ok/0004_string_value.graphql
@@ -2,9 +2,6 @@
 "simple"
 " white space "
 "unicode \u1234\u5678\u90AB\uCDEF"
-"string with unicode surrogate pair escape \uD83D\uDE00"
-"string with minimal surrogate pair escape \uD800\uDC00"
-"string with maximal surrogate pair escape \uDBFF\uDFFF"
 "string with \"escaped\" characters"
 "string with multiple languages котя, 猫, ねこ, قطة"
 """

diff --git a/crates/apollo-parser/test_data/lexer/ok/0004_string_value.txt b/crates/apollo-parser/test_data/lexer/ok/0004_string_value.txt
@@ -6,16 +6,10 @@ STRING_VALUE@12:27 "\" white space \""
 WHITESPACE@27:28 "\n"
 STRING_VALUE@28:62 "\"unicode \\u1234\\u5678\\u90AB\\uCDEF\""
 WHITESPACE@62:63 "\n"
-STRING_VALUE@63:119 "\"string with unicode surrogate pair escape \\uD83D\\uDE00\""
-WHITESPACE@119:120 "\n"
-STRING_VALUE@120:176 "\"string with minimal surrogate pair escape \\uD800\\uDC00\""
-WHITESPACE@176:177 "\n"
-STRING_VALUE@177:233 "\"string with maximal surrogate pair escape \\uDBFF\\uDFFF\""
-WHITESPACE@233:234 "\n"
-STRING_VALUE@234:270 "\"string with \\\"escaped\\\" characters\""
-WHITESPACE@270:271 "\n"
-STRING_VALUE@271:333 "\"string with multiple languages котя, 猫, ねこ, قطة\""
-WHITESPACE@333:334 "\n"
-STRING_VALUE@334:421 "\"\"\"\nblock string with unusual whitespaces\na b  c\nd\n\ne\tf\ng\u{2028}\u{2029}h\ni\u{b}j\u{c}k\u{feff}l\u{85}\u{200e}\u{200f}m\n\"\"\""
-WHITESPACE@421:422 "\n"
-EOF@422:422
+STRING_VALUE@63:99 "\"string with \\\"escaped\\\" characters\""
+WHITESPACE@99:100 "\n"
+STRING_VALUE@100:162 "\"string with multiple languages котя, 猫, ねこ, قطة\""
+WHITESPACE@162:163 "\n"
+STRING_VALUE@163:250 "\"\"\"\nblock string with unusual whitespaces\na b  c\nd\n\ne\tf\ng\u{2028}\u{2029}h\ni\u{b}j\u{c}k\u{feff}l\u{85}\u{200e}\u{200f}m\n\"\"\""
+WHITESPACE@250:251 "\n"
+EOF@251:251