Skip to content

Commit 11692f9

Browse files
pchampinpchampin
committed
split crate in 2
* statement contains only traits and simple wrapper types * statement_validation contains code for validating the contract of the wrapper types
1 parent acd3ba4 commit 11692f9

File tree

10 files changed

+580
-496
lines changed

10 files changed

+580
-496
lines changed

‎Cargo.toml

Lines changed: 4 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -3,6 +3,7 @@
33
members = [
44
"dummy",
55
"statement",
6+
"statement_validation",
67
]
78
resolver = "3"
89

@@ -16,7 +17,9 @@ license-file = "./LICENSE.md"
1617
keywords = ["rdf", "linked-data", "semantic-web", "w3c"] # no more than 5
1718

1819
[workspace.dependencies]
19-
dummy = { version = "0.1.0", path = "dummy" }
20+
r2c2_dummy = { version = "0.1.0", path = "dummy" }
21+
r2c2_statement = { version = "0.1.0", path = "statement" }
22+
r2c2_statement_validation = { version = "0.1.0", path = "statement_validation" }
2023

2124
[workspace.lints.clippy]
2225
enum_glob_use = "allow"

‎statement/Cargo.toml

Lines changed: 0 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -9,8 +9,6 @@ license-file.workspace = true
99
keywords.workspace = true
1010

1111
[dependencies]
12-
regex = "1.11.1"
13-
1412
langtag = { version = "0.4.0", optional = true }
1513
oxrdf = { version = "0.2.4", optional = true, features = ["rdf-star"] }
1614
rdf-types = { version = "0.22.5", optional = true }

‎statement/src/_iri.rs

Lines changed: 1 addition & 204 deletions
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,4 @@
1-
use std::{borrow::Cow, sync::LazyLock};
2-
3-
use regex::Regex;
1+
use std::borrow::Cow;
42

53
/// Wrapper around a [`Cow<str>`] guaranteeing that the underlying text satisfies [RFC3987].
64
///
@@ -9,12 +7,6 @@ use regex::Regex;
97
pub struct Iri<'a>(Cow<'a, str>);
108

119
impl<'a> Iri<'a> {
12-
/// Return a new [`Iri`] if the argument is a valid IRI, otherwise None.
13-
pub fn new(txt: impl Into<Cow<'a, str>>) -> Option<Self> {
14-
let inner = txt.into();
15-
IRI_REGEX.is_match(&inner).then_some(Iri(inner))
16-
}
17-
1810
/// Return a new [`Iri`], assuming the argument is a valid IRI.
1911
pub fn new_unchecked(txt: impl Into<Cow<'a, str>>) -> Self {
2012
Iri(txt.into())
@@ -86,142 +78,6 @@ impl std::fmt::Display for Iri<'_> {
8678
}
8779
}
8880

89-
pub(crate) static IRI_REGEX: LazyLock<Regex> = LazyLock::new(|| Regex::new(IRI_REGEX_SRC).unwrap());
90-
91-
/// Match an absolute IRI reference.
92-
pub static IRI_REGEX_SRC: &str = r"(?x)^
93-
#scheme
94-
( # CAPTURE scheme
95-
[A-Za-z] [-A-Za-z0-9+.]*
96-
)
97-
:
98-
#ihier_part
99-
(?: #iauthority + ipath_abempty
100-
//
101-
( # CAPTURE iauthority
102-
(?: # iuserinfo
103-
(?: [-A-Za-z0-9._~\u{A0}-\u{D7FF}\u{F900}-\u{FDCF}\u{FDF0}-\u{FFEF}\u{10000}-\u{1FFFD}\u{20000}-\u{2FFFD}\u{30000}-\u{3FFFD}\u{40000}-\u{4FFFD}\u{50000}-\u{5FFFD}\u{60000}-\u{6FFFD}\u{70000}-\u{7FFFD}\u{80000}-\u{8FFFD}\u{90000}-\u{9FFFD}\u{A0000}-\u{AFFFD}\u{B0000}-\u{BFFFD}\u{C0000}-\u{CFFFD}\u{D0000}-\u{DFFFD}\u{E1000}-\u{EFFFD}!$&'()*+,;=:]
104-
|
105-
%[0-9a-fA-F]{2}
106-
)*
107-
@
108-
)?
109-
# ihost
110-
(?: # ip_literal
111-
\[
112-
(?: # ipv6address
113-
(?:
114-
(?:[0-9a-fA-F]{1,4}:){6}
115-
(?:[0-9a-fA-F]{1,4}:[0-9a-fA-F]{1,4}|(?:[0-9]|(?:[1-9][0-9])|(?:1[0-9]{2})|(?:2[0-4][0-9])|(?:25[0-5]))(?:\.(?:[0-9]|(?:[1-9][0-9])|(?:1[0-9]{2})|(?:2[0-4][0-9])|(?:25[0-5]))){3})
116-
|
117-
::
118-
(?:[0-9a-fA-F]{1,4}:){5}
119-
(?:[0-9a-fA-F]{1,4}:[0-9a-fA-F]{1,4}|(?:[0-9]|(?:[1-9][0-9])|(?:1[0-9]{2})|(?:2[0-4][0-9])|(?:25[0-5]))(?:\.(?:[0-9]|(?:[1-9][0-9])|(?:1[0-9]{2})|(?:2[0-4][0-9])|(?:25[0-5]))){3})
120-
|
121-
(?:[0-9a-fA-F]{1,4})?
122-
::
123-
(?:[0-9a-fA-F]{1,4}:){4}
124-
(?:[0-9a-fA-F]{1,4}:[0-9a-fA-F]{1,4}|(?:[0-9]|(?:[1-9][0-9])|(?:1[0-9]{2})|(?:2[0-4][0-9])|(?:25[0-5]))(?:\.(?:[0-9]|(?:[1-9][0-9])|(?:1[0-9]{2})|(?:2[0-4][0-9])|(?:25[0-5]))){3})
125-
|
126-
(?:(?:[0-9a-fA-F]{1,4}:){0,1}:[0-9a-fA-F]{1,4})?
127-
::
128-
(?:[0-9a-fA-F]{1,4}:){3}
129-
(?:[0-9a-fA-F]{1,4}:[0-9a-fA-F]{1,4}|(?:[0-9]|(?:[1-9][0-9])|(?:1[0-9]{2})|(?:2[0-4][0-9])|(?:25[0-5]))(?:\.(?:[0-9]|(?:[1-9][0-9])|(?:1[0-9]{2})|(?:2[0-4][0-9])|(?:25[0-5]))){3})
130-
|
131-
(?:(?:[0-9a-fA-F]{1,4}:){0,2}:[0-9a-fA-F]{1,4})?
132-
::
133-
(?:[0-9a-fA-F]{1,4}:){2}
134-
(?:[0-9a-fA-F]{1,4}:[0-9a-fA-F]{1,4}|(?:[0-9]|(?:[1-9][0-9])|(?:1[0-9]{2})|(?:2[0-4][0-9])|(?:25[0-5]))(?:\.(?:[0-9]|(?:[1-9][0-9])|(?:1[0-9]{2})|(?:2[0-4][0-9])|(?:25[0-5]))){3})
135-
|
136-
(?:(?:[0-9a-fA-F]{1,4}:){0,3}:[0-9a-fA-F]{1,4})?
137-
::
138-
[0-9a-fA-F]{1,4}:
139-
(?:[0-9a-fA-F]{1,4}:[0-9a-fA-F]{1,4}|(?:[0-9]|(?:[1-9][0-9])|(?:1[0-9]{2})|(?:2[0-4][0-9])|(?:25[0-5]))(?:\.(?:[0-9]|(?:[1-9][0-9])|(?:1[0-9]{2})|(?:2[0-4][0-9])|(?:25[0-5]))){3})
140-
|
141-
(?:(?:[0-9a-fA-F]{1,4}:){0,4}:[0-9a-fA-F]{1,4})?
142-
::
143-
(?:[0-9a-fA-F]{1,4}:[0-9a-fA-F]{1,4}|(?:[0-9]|(?:[1-9][0-9])|(?:1[0-9]{2})|(?:2[0-4][0-9])|(?:25[0-5]))(?:\.(?:[0-9]|(?:[1-9][0-9])|(?:1[0-9]{2})|(?:2[0-4][0-9])|(?:25[0-5]))){3})
144-
|
145-
(?:(?:[0-9a-fA-F]{1,4}:){0,5}:[0-9a-fA-F]{1,4})?
146-
::
147-
[0-9a-fA-F]{1,4}
148-
|
149-
(?:(?:[0-9a-fA-F]{1,4}:){0,6}:[0-9a-fA-F]{1,4})?
150-
::
151-
)
152-
| # ipvfuture
153-
v[0-9a-fA-F]+ \. [-A-Za-z0-9._~!$&'()*+,;=:]+
154-
)
155-
\]
156-
| # ipv4address
157-
(?:[0-9]|(?:[1-9][0-9])|(?:1[0-9]{2})|(?:2[0-4][0-9])|(?:25[0-5])) (?:\.(?:[0-9]|(?:[1-9][0-9])|(?:1[0-9]{2})|(?:2[0-4][0-9])|(?:25[0-5]))){3}
158-
| # ireg_name
159-
(?: [-A-Za-z0-9._~\u{A0}-\u{D7FF}\u{F900}-\u{FDCF}\u{FDF0}-\u{FFEF}\u{10000}-\u{1FFFD}\u{20000}-\u{2FFFD}\u{30000}-\u{3FFFD}\u{40000}-\u{4FFFD}\u{50000}-\u{5FFFD}\u{60000}-\u{6FFFD}\u{70000}-\u{7FFFD}\u{80000}-\u{8FFFD}\u{90000}-\u{9FFFD}\u{A0000}-\u{AFFFD}\u{B0000}-\u{BFFFD}\u{C0000}-\u{CFFFD}\u{D0000}-\u{DFFFD}\u{E1000}-\u{EFFFD}!$&'()*+,;=]
160-
| %[0-9a-fA-F]{2}
161-
)*
162-
)
163-
(?:
164-
:
165-
[0-9]* # port
166-
)?
167-
)
168-
#ipath_abempty
169-
( # CAPTURE ipath_abempty
170-
(?:
171-
/
172-
(?: [-A-Za-z0-9._~\u{A0}-\u{D7FF}\u{F900}-\u{FDCF}\u{FDF0}-\u{FFEF}\u{10000}-\u{1FFFD}\u{20000}-\u{2FFFD}\u{30000}-\u{3FFFD}\u{40000}-\u{4FFFD}\u{50000}-\u{5FFFD}\u{60000}-\u{6FFFD}\u{70000}-\u{7FFFD}\u{80000}-\u{8FFFD}\u{90000}-\u{9FFFD}\u{A0000}-\u{AFFFD}\u{B0000}-\u{BFFFD}\u{C0000}-\u{CFFFD}\u{D0000}-\u{DFFFD}\u{E1000}-\u{EFFFD}!$&'()*+,;=:@]
173-
| %[0-9a-fA-F]{2}
174-
)*
175-
)*
176-
)
177-
| #ipath_absolute
178-
( # CAPTURE ipath_absolute
179-
/
180-
(?:
181-
(?: [-A-Za-z0-9._~\u{A0}-\u{D7FF}\u{F900}-\u{FDCF}\u{FDF0}-\u{FFEF}\u{10000}-\u{1FFFD}\u{20000}-\u{2FFFD}\u{30000}-\u{3FFFD}\u{40000}-\u{4FFFD}\u{50000}-\u{5FFFD}\u{60000}-\u{6FFFD}\u{70000}-\u{7FFFD}\u{80000}-\u{8FFFD}\u{90000}-\u{9FFFD}\u{A0000}-\u{AFFFD}\u{B0000}-\u{BFFFD}\u{C0000}-\u{CFFFD}\u{D0000}-\u{DFFFD}\u{E1000}-\u{EFFFD}!$&'()*+,;=:@]
182-
| %[0-9a-fA-F]{2}
183-
)*
184-
(?:
185-
/
186-
(?: [-A-Za-z0-9._~\u{A0}-\u{D7FF}\u{F900}-\u{FDCF}\u{FDF0}-\u{FFEF}\u{10000}-\u{1FFFD}\u{20000}-\u{2FFFD}\u{30000}-\u{3FFFD}\u{40000}-\u{4FFFD}\u{50000}-\u{5FFFD}\u{60000}-\u{6FFFD}\u{70000}-\u{7FFFD}\u{80000}-\u{8FFFD}\u{90000}-\u{9FFFD}\u{A0000}-\u{AFFFD}\u{B0000}-\u{BFFFD}\u{C0000}-\u{CFFFD}\u{D0000}-\u{DFFFD}\u{E1000}-\u{EFFFD}!$&'()*+,;=:@]
187-
| %[0-9a-fA-F]{2}
188-
)*
189-
)*
190-
)?
191-
)
192-
| #ipath_rootless
193-
( # CAPTURE ipath_rootless
194-
(?: [-A-Za-z0-9._~\u{A0}-\u{D7FF}\u{F900}-\u{FDCF}\u{FDF0}-\u{FFEF}\u{10000}-\u{1FFFD}\u{20000}-\u{2FFFD}\u{30000}-\u{3FFFD}\u{40000}-\u{4FFFD}\u{50000}-\u{5FFFD}\u{60000}-\u{6FFFD}\u{70000}-\u{7FFFD}\u{80000}-\u{8FFFD}\u{90000}-\u{9FFFD}\u{A0000}-\u{AFFFD}\u{B0000}-\u{BFFFD}\u{C0000}-\u{CFFFD}\u{D0000}-\u{DFFFD}\u{E1000}-\u{EFFFD}!$&'()*+,;=:@]
195-
| %[0-9a-fA-F]{2}
196-
)+
197-
(?:
198-
/
199-
(?: [-A-Za-z0-9._~\u{A0}-\u{D7FF}\u{F900}-\u{FDCF}\u{FDF0}-\u{FFEF}\u{10000}-\u{1FFFD}\u{20000}-\u{2FFFD}\u{30000}-\u{3FFFD}\u{40000}-\u{4FFFD}\u{50000}-\u{5FFFD}\u{60000}-\u{6FFFD}\u{70000}-\u{7FFFD}\u{80000}-\u{8FFFD}\u{90000}-\u{9FFFD}\u{A0000}-\u{AFFFD}\u{B0000}-\u{BFFFD}\u{C0000}-\u{CFFFD}\u{D0000}-\u{DFFFD}\u{E1000}-\u{EFFFD}!$&'()*+,;=:@]
200-
| %[0-9a-fA-F]{2}
201-
)*
202-
)*
203-
)
204-
)? # optional because of ipath_empty
205-
(?: # ?iquery
206-
\?
207-
( # CAPTURE iquery
208-
(?:
209-
[-A-Za-z0-9._~\u{A0}-\u{D7FF}\u{F900}-\u{FDCF}\u{FDF0}-\u{FFEF}\u{10000}-\u{1FFFD}\u{20000}-\u{2FFFD}\u{30000}-\u{3FFFD}\u{40000}-\u{4FFFD}\u{50000}-\u{5FFFD}\u{60000}-\u{6FFFD}\u{70000}-\u{7FFFD}\u{80000}-\u{8FFFD}\u{90000}-\u{9FFFD}\u{A0000}-\u{AFFFD}\u{B0000}-\u{BFFFD}\u{C0000}-\u{CFFFD}\u{D0000}-\u{DFFFD}\u{E1000}-\u{EFFFD}!$&'()*+,;=:@'\u{E000}-\u{F8FF}\u{F0000}-\u{FFFFD}\u{100000}-\u{10FFFD}/?]
210-
| %[0-9a-fA-F]{2}
211-
)*
212-
)
213-
)?
214-
(?: # #ifragment
215-
\#
216-
( # CAPTURE ifragment
217-
(?:
218-
[-A-Za-z0-9._~\u{A0}-\u{D7FF}\u{F900}-\u{FDCF}\u{FDF0}-\u{FFEF}\u{10000}-\u{1FFFD}\u{20000}-\u{2FFFD}\u{30000}-\u{3FFFD}\u{40000}-\u{4FFFD}\u{50000}-\u{5FFFD}\u{60000}-\u{6FFFD}\u{70000}-\u{7FFFD}\u{80000}-\u{8FFFD}\u{90000}-\u{9FFFD}\u{A0000}-\u{AFFFD}\u{B0000}-\u{BFFFD}\u{C0000}-\u{CFFFD}\u{D0000}-\u{DFFFD}\u{E1000}-\u{EFFFD}!$&'()*+,;=:@/?]
219-
| %[0-9a-fA-F]{2}
220-
)*
221-
)
222-
)?
223-
$";
224-
22581
#[cfg(test)]
22682
mod test {
22783
use super::*;
@@ -250,63 +106,4 @@ mod test {
250106
let iri1 = Iri::new_unchecked(ex.to_string());
251107
assert_eq!(iri1.to_string(), format!("<{ex}>"));
252108
}
253-
254-
#[test]
255-
fn regex() {
256-
for txt in POSITIVE_IRIS {
257-
assert!(IRI_REGEX.is_match(txt));
258-
}
259-
for txt in NEGATIVE_IRIS {
260-
assert!(!IRI_REGEX.is_match(txt));
261-
}
262-
}
263-
264-
/// An array of valid IRIs
265-
pub const POSITIVE_IRIS: &[&str] = &[
266-
"http:",
267-
"http://example.org",
268-
"http://127.0.0.1",
269-
"http://[::]",
270-
"http://%0D",
271-
"http://example.org/",
272-
"http://éxample.org/",
273-
"http://user:[email protected]:1234/",
274-
"http://example.org/foo/bar/baz",
275-
"http://example.org/foo/bar/",
276-
"http://example.org/foo/bar/bàz",
277-
"http://example.org/foo/.././/bar",
278-
"http://example.org/!$&'()*+,=:@/foo%0D",
279-
"http://example.org/?abc",
280-
"http://example.org/?!$&'()*+,=:@/?\u{E000}",
281-
"http://example.org/#def",
282-
"http://example.org/?abc#def",
283-
"tag:abc/def",
284-
"tag:",
285-
"http://example.org/#Andr%C3%A9",
286-
"http://example.org/?Andr%C3%A9",
287-
];
288-
289-
/// An array of invalid IRIs.
290-
pub const NEGATIVE_IRIS: &[&str] = &[
291-
// valid IRI references that are not IRIs (relative)
292-
"foo",
293-
"..",
294-
"//example.org",
295-
"?",
296-
"#",
297-
"?#",
298-
"?Andr%C3%A9#Andr%C3%A9",
299-
// invalid IRI references
300-
"http://[/",
301-
"http://a/[",
302-
"http://a/]",
303-
"http://a/|",
304-
"http://a/ ",
305-
"http://a/\u{E000}",
306-
"[",
307-
"]",
308-
"|",
309-
" ",
310-
"\u{E000}",
311-
];
312109
}

0 commit comments

Comments
 (0)