1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
use super::types::request::*;
use crate::utils::{self, MatchedFiles};
use getset::Getters;
use regex::Regex;
#[cfg(feature = "reqwest")]
use reqwest::header::HeaderMap;
use std::time::Duration;
mod macros;
pub use gemini_proc_macros::{
execute_function_calls, execute_function_calls_with_callback, gemini_function, gemini_schema,
};
pub use macros::GeminiSchema;
const REQ_TIMEOUT: Duration = Duration::from_secs(10);
pub struct MarkdownToPartsBuilder {
regex: Option<Regex>,
guess_mime_type: Option<fn(url: &str) -> mime::Mime>,
#[cfg(feature = "reqwest")]
decide_download: Option<fn(headers: &HeaderMap) -> bool>,
timeout: Option<Duration>,
}
impl MarkdownToPartsBuilder {
///# Panics
///`regex` must have a Regex with only 1 capture group with file URL as first capture
///group, else it PANICS when `.build()` is called.
pub fn regex(mut self, regex: Regex) -> Self {
self.regex = Some(regex);
self
}
/// `guess_mime_type` is used to detect mimi_type of URL pointing to file system or web resource
/// with no "Content-Type" header.
pub fn guess_mime_type(mut self, guess_mime_type: fn(url: &str) -> mime::Mime) -> Self {
self.guess_mime_type = Some(guess_mime_type);
self
}
/// `decide_download` is used to decide if to download. If it returns false, resource will not
/// be fetched and won't be in `parts`
#[cfg(feature = "reqwest")]
pub fn decide_download(mut self, decide_download: fn(headers: &HeaderMap) -> bool) -> Self {
self.decide_download = Some(decide_download);
self
}
pub fn timeout(mut self, timeout: Duration) -> Self {
self.timeout = Some(timeout);
self
}
#[cfg(feature = "reqwest")]
pub async fn build<'a>(self, markdown: &'a str) -> MarkdownToParts<'a> {
MarkdownToParts {
markdown,
base64s: utils::get_file_base64s(
markdown,
self.regex
.unwrap_or(Regex::new(r"(?s)!\[.*?].?\((.*?)\)").unwrap()),
self.guess_mime_type.unwrap_or(|_| mime::IMAGE_PNG),
self.decide_download.unwrap_or(|_| true),
self.timeout.unwrap_or(REQ_TIMEOUT),
)
.await,
}
}
}
#[derive(Getters, Clone)]
///Converts markdown to parts considering `` means Gemini will be see the images too. `link` can be URL or file path.
pub struct MarkdownToParts<'a> {
markdown: &'a str,
#[get = "pub"]
base64s: Vec<MatchedFiles>,
}
impl<'a> MarkdownToParts<'a> {
pub fn builder() -> MarkdownToPartsBuilder {
MarkdownToPartsBuilder {
regex: None,
guess_mime_type: None,
#[cfg(feature = "reqwest")]
decide_download: None,
timeout: None,
}
}
///# Panics
///`regex` must have a Regex with only 1 capture group with file URL as first capture
///group, else it PANICS.
/// # Arguments
/// `guess_mime_type` is used to detect mimi_type of URL pointing to file system or web resource
/// with no "Content-Type" header.
/// `decide_download` is used to decide if to download. If it returns false, resource will not
/// be fetched and won't be in `parts`
/// # Example
/// ```ignore
/// from_regex("Your markdown string...", Regex::new(r"(?s)!\[.*?].?\((.*?)\)").unwrap(), |_| mime::IMAGE_PNG, |_| true)
/// ```
#[cfg(feature = "reqwest")]
pub async fn from_regex_checked(
markdown: &'a str,
regex: Regex,
guess_mime_type: fn(url: &str) -> mime::Mime,
decide_download: fn(headers: &HeaderMap) -> bool,
) -> Self {
Self {
base64s: utils::get_file_base64s(
markdown,
regex,
guess_mime_type,
decide_download,
REQ_TIMEOUT,
)
.await,
markdown,
}
}
///# Panics
///`regex` must have a Regex with only 1 capture group with file URL as first capture
///group, else it PANICS.
/// # Arguments
/// `guess_mime_type` is used to detect mimi_type of URL pointing to file system or web resource
/// with no "Content-Type" header.
/// # Example
/// ```ignore
/// from_regex("Your markdown string...", Regex::new(r"(?s)!\[.*?].?\((.*?)\)").unwrap(), |_|
/// mime::IMAGE_PNG)
/// ```
#[cfg(feature = "reqwest")]
pub async fn from_regex(
markdown: &'a str,
regex: Regex,
guess_mime_type: fn(url: &str) -> mime::Mime,
) -> Self {
Self::from_regex_checked(markdown, regex, guess_mime_type, |_| true).await
}
/// # Arguments
/// `guess_mime_type` is used to detect mimi_type of URL pointing to file system or web resource
/// with no "Content-Type" header.
/// `decide_download` is used to decide if to download. If it returns false, resource will not
/// be fetched and won't be in `parts`
/// # Example
/// ```ignore
/// new("Your markdown string...", |_| mime::IMAGE_PNG, |_| true)
/// ```
#[cfg(feature = "reqwest")]
pub async fn new_checked(
markdown: &'a str,
guess_mime_type: fn(url: &str) -> mime::Mime,
decide_download: fn(headers: &HeaderMap) -> bool,
) -> Self {
let image_regex = Regex::new(r"(?s)!\[.*?].?\((.*?)\)").unwrap();
Self::from_regex_checked(markdown, image_regex, guess_mime_type, decide_download).await
}
/// # Arguments
/// `guess_mime_type` is used to detect mimi_type of URL pointing to file system or web resource
/// with no "Content-Type" header.
/// # Example
/// ```ignore
/// new("Your markdown string...", |_| mime::IMAGE_PNG)
/// ```
#[cfg(feature = "reqwest")]
pub async fn new(markdown: &'a str, guess_mime_type: fn(url: &str) -> mime::Mime) -> Self {
Self::new_checked(markdown, guess_mime_type, |_| true).await
}
pub fn process(mut self) -> Vec<Part> {
let mut parts: Vec<Part> = Vec::new();
let mut removed_length = 0;
for file in self.base64s {
if let MatchedFiles {
index,
length,
mime_type: Some(mime_type),
base64: Some(base64),
} = file
{
let end = index + length - removed_length;
let text = &self.markdown[..end];
parts.push(text.into());
parts.push(InlineData::new(mime_type, base64).into());
self.markdown = &self.markdown[end..];
removed_length += end;
}
}
if self.markdown.len() != 0 {
parts.push(self.markdown.into());
}
parts
}
}