1
//! String-manipulation utilities
2

            
3
/// Return the position of one string slice within another.
4
///
5
/// If `needle` is indeed part of `haystack`, returns some offset
6
/// `off`, such that `needle` is the same as
7
/// `&haystack[off..needle.len()]`.
8
///
9
/// Returns None if `needle` is not a part of `haystack`.
10
///
11
/// Remember, offsets are in bytes, not in characters.
12
///
13
/// # Example
14
/// ```ignore
15
/// use tor_netdoc::util::str_offset;
16
/// let quote = "A rose is a rose is a rose."; // -- Gertrude Stein
17
/// assert_eq!(&quote[2..6], "rose");
18
/// assert_eq!(str_offset(quote, &quote[2..6]).unwrap(), 2);
19
/// assert_eq!(&quote[12..16], "rose");
20
/// assert_eq!(str_offset(quote, &quote[12..16]).unwrap(), 12);
21
/// assert_eq!(&quote[22..26], "rose");
22
/// assert_eq!(str_offset(quote, &quote[22..26]).unwrap(), 22);
23
///
24
/// assert_eq!(str_offset(quote, "rose"), None);
25
///
26
/// assert_eq!(str_offset(&quote[1..], &quote[2..6]), Some(1));
27
/// assert_eq!(str_offset(&quote[1..5], &quote[2..6]), None);
28
/// ```
29
8711
pub(crate) fn str_offset(haystack: &str, needle: &str) -> Option<usize> {
30
8711
    let needle_start_u = needle.as_ptr() as usize;
31
8711
    let needle_end_u = needle_start_u + needle.len();
32
8711
    let haystack_start_u = haystack.as_ptr() as usize;
33
8711
    let haystack_end_u = haystack_start_u + haystack.len();
34
8711
    if haystack_start_u <= needle_start_u && needle_end_u <= haystack_end_u {
35
8705
        Some(needle_start_u - haystack_start_u)
36
    } else {
37
6
        None
38
    }
39
8711
}
40

            
41
/// An extent within a given string slice.
42
///
43
/// This whole type is probably naughty and shouldn't exist.  We use
44
/// it only within this crate, to remember where it was that we found
45
/// parsed objects within the strings we got them from.
46
#[derive(Clone, Debug)]
47
pub(crate) struct Extent {
48
    /// At what position within the original string is this extent, in bytes?
49
    offset: usize,
50
    /// How long is this extend, in bytes?
51
    length: usize,
52
    /// What was the original string?
53
    ///
54
    /// If this doesn't match, there's been an error.
55
    sliceptr: *const u8,
56
    /// How long was the original string?
57
    ///
58
    /// If this doesn't match, there's been an error.
59
    slicelen: usize,
60
}
61

            
62
impl Extent {
63
    /// Construct a new extent to represent the position of `needle`
64
    /// within `haystack`.
65
    ///
66
    /// Return None if `needle` is not in fact a slice of `haystack`.
67
489
    pub(crate) fn new(haystack: &str, needle: &str) -> Option<Extent> {
68
518
        str_offset(haystack, needle).map(|offset| Extent {
69
487
            offset,
70
487
            length: needle.len(),
71
487
            sliceptr: haystack.as_ptr(),
72
487
            slicelen: haystack.len(),
73
518
        })
74
489
    }
75
    /// Reconstruct the original `needle` within `haystack`.
76
    ///
77
    /// Return None if we're sure that the haystack doesn't match the one
78
    /// we were originally given.
79
    ///
80
    /// Note that it is possible for this to give a bogus result if
81
    /// provided a new haystack that happens to be at the same
82
    /// position in memory as the original one.
83
414
    pub(crate) fn reconstruct<'a>(&self, haystack: &'a str) -> Option<&'a str> {
84
414
        if self.sliceptr != haystack.as_ptr() || self.slicelen != haystack.len() {
85
2
            None
86
        } else {
87
412
            haystack.get(self.offset..self.offset + self.length)
88
        }
89
414
    }
90
}
91

            
92
#[cfg(test)]
93
mod test {
94
    // @@ begin test lint list maintained by maint/add_warning @@
95
    #![allow(clippy::bool_assert_comparison)]
96
    #![allow(clippy::clone_on_copy)]
97
    #![allow(clippy::dbg_macro)]
98
    #![allow(clippy::mixed_attributes_style)]
99
    #![allow(clippy::print_stderr)]
100
    #![allow(clippy::print_stdout)]
101
    #![allow(clippy::single_char_pattern)]
102
    #![allow(clippy::unwrap_used)]
103
    #![allow(clippy::unchecked_duration_subtraction)]
104
    #![allow(clippy::useless_vec)]
105
    #![allow(clippy::needless_pass_by_value)]
106
    //! <!-- @@ end test lint list maintained by maint/add_warning @@ -->
107

            
108
    #[test]
109
    fn test_str_offset() {
110
        use super::str_offset;
111
        let quote = "A rose is a rose is a rose."; // -- Gertrude Stein
112
        assert_eq!(&quote[2..6], "rose");
113
        assert_eq!(str_offset(quote, &quote[2..6]).unwrap(), 2);
114
        assert_eq!(&quote[12..16], "rose");
115
        assert_eq!(str_offset(quote, &quote[12..16]).unwrap(), 12);
116
        assert_eq!(&quote[22..26], "rose");
117
        assert_eq!(str_offset(quote, &quote[22..26]).unwrap(), 22);
118

            
119
        assert_eq!(str_offset(quote, "rose"), None);
120

            
121
        assert_eq!(str_offset(&quote[1..], &quote[2..6]), Some(1));
122
        assert_eq!(str_offset(&quote[1..5], &quote[2..6]), None);
123
    }
124

            
125
    #[test]
126
    fn test_extent() {
127
        use super::Extent;
128
        let quote = "What is a winter wedding a winter wedding."; // -- ibid
129
        assert_eq!(&quote[10..16], "winter");
130
        let ex = Extent::new(quote, &quote[10..16]).unwrap();
131
        let s = ex.reconstruct(quote).unwrap();
132
        assert_eq!(s, "winter");
133

            
134
        assert!(Extent::new(quote, "winter").is_none());
135
        assert!(ex.reconstruct("Hello world").is_none());
136
    }
137
}