297 lines
28 KiB
HTML
297 lines
28 KiB
HTML
<!DOCTYPE html><html lang="en"><head><meta charset="utf-8"><meta name="viewport" content="width=device-width, initial-scale=1.0"><meta name="generator" content="rustdoc"><meta name="description" content="Source of the Rust file `/home/newkirk/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/zstd-0.13.3/src/dict.rs`."><title>dict.rs - source</title><script>if(window.location.protocol!=="file:")document.head.insertAdjacentHTML("beforeend","SourceSerif4-Regular-6b053e98.ttf.woff2,FiraSans-Italic-81dc35de.woff2,FiraSans-Regular-0fe48ade.woff2,FiraSans-MediumItalic-ccf7e434.woff2,FiraSans-Medium-e1aa3f0a.woff2,SourceCodePro-Regular-8badfe75.ttf.woff2,SourceCodePro-Semibold-aa29a496.ttf.woff2".split(",").map(f=>`<link rel="preload" as="font" type="font/woff2"href="../../static.files/${f}">`).join(""))</script><link rel="stylesheet" href="../../static.files/normalize-9960930a.css"><link rel="stylesheet" href="../../static.files/rustdoc-ca0dd0c4.css"><meta name="rustdoc-vars" data-root-path="../../" data-static-root-path="../../static.files/" data-current-crate="zstd" data-themes="" data-resource-suffix="" data-rustdoc-version="1.93.1 (01f6ddf75 2026-02-11) (Arch Linux rust 1:1.93.1-1)" data-channel="1.93.1" data-search-js="search-9e2438ea.js" data-stringdex-js="stringdex-a3946164.js" data-settings-js="settings-c38705f0.js" ><script src="../../static.files/storage-e2aeef58.js"></script><script defer src="../../static.files/src-script-813739b1.js"></script><script defer src="../../src-files.js"></script><script defer src="../../static.files/main-a410ff4d.js"></script><noscript><link rel="stylesheet" href="../../static.files/noscript-263c88ec.css"></noscript><link rel="alternate icon" type="image/png" href="../../static.files/favicon-32x32-eab170b8.png"><link rel="icon" type="image/svg+xml" href="../../static.files/favicon-044be391.svg"></head><body class="rustdoc src"><!--[if lte IE 11]><div class="warning">This old browser is unsupported and will most likely display funky things.</div><![endif]--><nav class="sidebar"><div class="src-sidebar-title"><h2>Files</h2></div></nav><div class="sidebar-resizer" title="Drag to resize sidebar"></div><main><section id="main-content" class="content"><div class="main-heading"><h1><div class="sub-heading">zstd/</div>dict.rs</h1><rustdoc-toolbar></rustdoc-toolbar></div><div class="example-wrap digits-3"><pre class="rust"><code><a href=#1 id=1 data-nosnippet>1</a><span class="doccomment">//! Train a dictionary from various sources.
|
|
<a href=#2 id=2 data-nosnippet>2</a>//!
|
|
<a href=#3 id=3 data-nosnippet>3</a>//! A dictionary can help improve the compression of small files.
|
|
<a href=#4 id=4 data-nosnippet>4</a>//! The dictionary must be present during decompression,
|
|
<a href=#5 id=5 data-nosnippet>5</a>//! but can be shared across multiple "similar" files.
|
|
<a href=#6 id=6 data-nosnippet>6</a>//!
|
|
<a href=#7 id=7 data-nosnippet>7</a>//! Creating a dictionary using the `zstd` C library,
|
|
<a href=#8 id=8 data-nosnippet>8</a>//! using the `zstd` command-line interface, using this library,
|
|
<a href=#9 id=9 data-nosnippet>9</a>//! or using the `train` binary provided, should give the same result,
|
|
<a href=#10 id=10 data-nosnippet>10</a>//! and are therefore completely compatible.
|
|
<a href=#11 id=11 data-nosnippet>11</a>//!
|
|
<a href=#12 id=12 data-nosnippet>12</a>//! To use, see [`Encoder::with_dictionary`] or [`Decoder::with_dictionary`].
|
|
<a href=#13 id=13 data-nosnippet>13</a>//!
|
|
<a href=#14 id=14 data-nosnippet>14</a>//! [`Encoder::with_dictionary`]: ../struct.Encoder.html#method.with_dictionary
|
|
<a href=#15 id=15 data-nosnippet>15</a>//! [`Decoder::with_dictionary`]: ../struct.Decoder.html#method.with_dictionary
|
|
<a href=#16 id=16 data-nosnippet>16</a>
|
|
<a href=#17 id=17 data-nosnippet>17</a></span><span class="attr">#[cfg(feature = <span class="string">"zdict_builder"</span>)]
|
|
<a href=#18 id=18 data-nosnippet>18</a></span><span class="kw">use </span>std::io::{<span class="self">self</span>, Read};
|
|
<a href=#19 id=19 data-nosnippet>19</a>
|
|
<a href=#20 id=20 data-nosnippet>20</a><span class="kw">pub use </span>zstd_safe::{CDict, DDict};
|
|
<a href=#21 id=21 data-nosnippet>21</a>
|
|
<a href=#22 id=22 data-nosnippet>22</a><span class="doccomment">/// Prepared dictionary for compression
|
|
<a href=#23 id=23 data-nosnippet>23</a>///
|
|
<a href=#24 id=24 data-nosnippet>24</a>/// A dictionary can include its own copy of the data (if it is `'static`), or it can merely point
|
|
<a href=#25 id=25 data-nosnippet>25</a>/// to a separate buffer (if it has another lifetime).
|
|
<a href=#26 id=26 data-nosnippet>26</a></span><span class="kw">pub struct </span>EncoderDictionary<<span class="lifetime">'a</span>> {
|
|
<a href=#27 id=27 data-nosnippet>27</a> cdict: CDict<<span class="lifetime">'a</span>>,
|
|
<a href=#28 id=28 data-nosnippet>28</a>}
|
|
<a href=#29 id=29 data-nosnippet>29</a>
|
|
<a href=#30 id=30 data-nosnippet>30</a><span class="kw">impl </span>EncoderDictionary<<span class="lifetime">'static</span>> {
|
|
<a href=#31 id=31 data-nosnippet>31</a> <span class="doccomment">/// Creates a prepared dictionary for compression.
|
|
<a href=#32 id=32 data-nosnippet>32</a> ///
|
|
<a href=#33 id=33 data-nosnippet>33</a> /// This will copy the dictionary internally.
|
|
<a href=#34 id=34 data-nosnippet>34</a> </span><span class="kw">pub fn </span>copy(dictionary: <span class="kw-2">&</span>[u8], level: i32) -> <span class="self">Self </span>{
|
|
<a href=#35 id=35 data-nosnippet>35</a> <span class="self">Self </span>{
|
|
<a href=#36 id=36 data-nosnippet>36</a> cdict: zstd_safe::create_cdict(dictionary, level),
|
|
<a href=#37 id=37 data-nosnippet>37</a> }
|
|
<a href=#38 id=38 data-nosnippet>38</a> }
|
|
<a href=#39 id=39 data-nosnippet>39</a>}
|
|
<a href=#40 id=40 data-nosnippet>40</a>
|
|
<a href=#41 id=41 data-nosnippet>41</a><span class="kw">impl</span><<span class="lifetime">'a</span>> EncoderDictionary<<span class="lifetime">'a</span>> {
|
|
<a href=#42 id=42 data-nosnippet>42</a> <span class="attr">#[cfg(feature = <span class="string">"experimental"</span>)]
|
|
<a href=#43 id=43 data-nosnippet>43</a> #[cfg_attr(feature = <span class="string">"doc-cfg"</span>, doc(cfg(feature = <span class="string">"experimental"</span>)))]
|
|
<a href=#44 id=44 data-nosnippet>44</a> </span><span class="doccomment">/// Create prepared dictionary for compression
|
|
<a href=#45 id=45 data-nosnippet>45</a> ///
|
|
<a href=#46 id=46 data-nosnippet>46</a> /// A level of `0` uses zstd's default (currently `3`).
|
|
<a href=#47 id=47 data-nosnippet>47</a> ///
|
|
<a href=#48 id=48 data-nosnippet>48</a> /// Only available with the `experimental` feature. Use `EncoderDictionary::copy` otherwise.
|
|
<a href=#49 id=49 data-nosnippet>49</a> </span><span class="kw">pub fn </span>new(dictionary: <span class="kw-2">&</span><span class="lifetime">'a </span>[u8], level: i32) -> <span class="self">Self </span>{
|
|
<a href=#50 id=50 data-nosnippet>50</a> <span class="self">Self </span>{
|
|
<a href=#51 id=51 data-nosnippet>51</a> cdict: zstd_safe::CDict::create_by_reference(dictionary, level),
|
|
<a href=#52 id=52 data-nosnippet>52</a> }
|
|
<a href=#53 id=53 data-nosnippet>53</a> }
|
|
<a href=#54 id=54 data-nosnippet>54</a>
|
|
<a href=#55 id=55 data-nosnippet>55</a> <span class="doccomment">/// Returns reference to `CDict` inner object
|
|
<a href=#56 id=56 data-nosnippet>56</a> </span><span class="kw">pub fn </span>as_cdict(<span class="kw-2">&</span><span class="self">self</span>) -> <span class="kw-2">&</span>CDict<<span class="lifetime">'a</span>> {
|
|
<a href=#57 id=57 data-nosnippet>57</a> <span class="kw-2">&</span><span class="self">self</span>.cdict
|
|
<a href=#58 id=58 data-nosnippet>58</a> }
|
|
<a href=#59 id=59 data-nosnippet>59</a>}
|
|
<a href=#60 id=60 data-nosnippet>60</a>
|
|
<a href=#61 id=61 data-nosnippet>61</a><span class="doccomment">/// Prepared dictionary for decompression
|
|
<a href=#62 id=62 data-nosnippet>62</a></span><span class="kw">pub struct </span>DecoderDictionary<<span class="lifetime">'a</span>> {
|
|
<a href=#63 id=63 data-nosnippet>63</a> ddict: DDict<<span class="lifetime">'a</span>>,
|
|
<a href=#64 id=64 data-nosnippet>64</a>}
|
|
<a href=#65 id=65 data-nosnippet>65</a>
|
|
<a href=#66 id=66 data-nosnippet>66</a><span class="kw">impl </span>DecoderDictionary<<span class="lifetime">'static</span>> {
|
|
<a href=#67 id=67 data-nosnippet>67</a> <span class="doccomment">/// Create a prepared dictionary for decompression.
|
|
<a href=#68 id=68 data-nosnippet>68</a> ///
|
|
<a href=#69 id=69 data-nosnippet>69</a> /// This will copy the dictionary internally.
|
|
<a href=#70 id=70 data-nosnippet>70</a> </span><span class="kw">pub fn </span>copy(dictionary: <span class="kw-2">&</span>[u8]) -> <span class="self">Self </span>{
|
|
<a href=#71 id=71 data-nosnippet>71</a> <span class="self">Self </span>{
|
|
<a href=#72 id=72 data-nosnippet>72</a> ddict: zstd_safe::DDict::create(dictionary),
|
|
<a href=#73 id=73 data-nosnippet>73</a> }
|
|
<a href=#74 id=74 data-nosnippet>74</a> }
|
|
<a href=#75 id=75 data-nosnippet>75</a>}
|
|
<a href=#76 id=76 data-nosnippet>76</a>
|
|
<a href=#77 id=77 data-nosnippet>77</a><span class="kw">impl</span><<span class="lifetime">'a</span>> DecoderDictionary<<span class="lifetime">'a</span>> {
|
|
<a href=#78 id=78 data-nosnippet>78</a> <span class="attr">#[cfg(feature = <span class="string">"experimental"</span>)]
|
|
<a href=#79 id=79 data-nosnippet>79</a> #[cfg_attr(feature = <span class="string">"doc-cfg"</span>, doc(cfg(feature = <span class="string">"experimental"</span>)))]
|
|
<a href=#80 id=80 data-nosnippet>80</a> </span><span class="doccomment">/// Create prepared dictionary for decompression
|
|
<a href=#81 id=81 data-nosnippet>81</a> ///
|
|
<a href=#82 id=82 data-nosnippet>82</a> /// Only available with the `experimental` feature. Use `DecoderDictionary::copy` otherwise.
|
|
<a href=#83 id=83 data-nosnippet>83</a> </span><span class="kw">pub fn </span>new(dict: <span class="kw-2">&</span><span class="lifetime">'a </span>[u8]) -> <span class="self">Self </span>{
|
|
<a href=#84 id=84 data-nosnippet>84</a> <span class="self">Self </span>{
|
|
<a href=#85 id=85 data-nosnippet>85</a> ddict: zstd_safe::DDict::create_by_reference(dict),
|
|
<a href=#86 id=86 data-nosnippet>86</a> }
|
|
<a href=#87 id=87 data-nosnippet>87</a> }
|
|
<a href=#88 id=88 data-nosnippet>88</a>
|
|
<a href=#89 id=89 data-nosnippet>89</a> <span class="doccomment">/// Returns reference to `DDict` inner object
|
|
<a href=#90 id=90 data-nosnippet>90</a> </span><span class="kw">pub fn </span>as_ddict(<span class="kw-2">&</span><span class="self">self</span>) -> <span class="kw-2">&</span>DDict<<span class="lifetime">'a</span>> {
|
|
<a href=#91 id=91 data-nosnippet>91</a> <span class="kw-2">&</span><span class="self">self</span>.ddict
|
|
<a href=#92 id=92 data-nosnippet>92</a> }
|
|
<a href=#93 id=93 data-nosnippet>93</a>}
|
|
<a href=#94 id=94 data-nosnippet>94</a>
|
|
<a href=#95 id=95 data-nosnippet>95</a><span class="doccomment">/// Train a dictionary from a big continuous chunk of data, with all samples
|
|
<a href=#96 id=96 data-nosnippet>96</a>/// contiguous in memory.
|
|
<a href=#97 id=97 data-nosnippet>97</a>///
|
|
<a href=#98 id=98 data-nosnippet>98</a>/// This is the most efficient way to train a dictionary,
|
|
<a href=#99 id=99 data-nosnippet>99</a>/// since this is directly fed into `zstd`.
|
|
<a href=#100 id=100 data-nosnippet>100</a>///
|
|
<a href=#101 id=101 data-nosnippet>101</a>/// * `sample_data` is the concatenation of all sample data.
|
|
<a href=#102 id=102 data-nosnippet>102</a>/// * `sample_sizes` is the size of each sample in `sample_data`.
|
|
<a href=#103 id=103 data-nosnippet>103</a>/// The sum of all `sample_sizes` should equal the length of `sample_data`.
|
|
<a href=#104 id=104 data-nosnippet>104</a>/// * `max_size` is the maximum size of the dictionary to generate.
|
|
<a href=#105 id=105 data-nosnippet>105</a>///
|
|
<a href=#106 id=106 data-nosnippet>106</a>/// The result is the dictionary data. You can, for example, feed it to [`CDict::create`].
|
|
<a href=#107 id=107 data-nosnippet>107</a></span><span class="attr">#[cfg(feature = <span class="string">"zdict_builder"</span>)]
|
|
<a href=#108 id=108 data-nosnippet>108</a>#[cfg_attr(feature = <span class="string">"doc-cfg"</span>, doc(cfg(feature = <span class="string">"zdict_builder"</span>)))]
|
|
<a href=#109 id=109 data-nosnippet>109</a></span><span class="kw">pub fn </span>from_continuous(
|
|
<a href=#110 id=110 data-nosnippet>110</a> sample_data: <span class="kw-2">&</span>[u8],
|
|
<a href=#111 id=111 data-nosnippet>111</a> sample_sizes: <span class="kw-2">&</span>[usize],
|
|
<a href=#112 id=112 data-nosnippet>112</a> max_size: usize,
|
|
<a href=#113 id=113 data-nosnippet>113</a>) -> io::Result<Vec<u8>> {
|
|
<a href=#114 id=114 data-nosnippet>114</a> <span class="kw">use </span><span class="kw">crate</span>::map_error_code;
|
|
<a href=#115 id=115 data-nosnippet>115</a>
|
|
<a href=#116 id=116 data-nosnippet>116</a> <span class="comment">// Complain if the lengths don't add up to the entire data.
|
|
<a href=#117 id=117 data-nosnippet>117</a> </span><span class="kw">if </span>sample_sizes.iter().sum::<usize>() != sample_data.len() {
|
|
<a href=#118 id=118 data-nosnippet>118</a> <span class="kw">return </span><span class="prelude-val">Err</span>(io::Error::new(
|
|
<a href=#119 id=119 data-nosnippet>119</a> io::ErrorKind::Other,
|
|
<a href=#120 id=120 data-nosnippet>120</a> <span class="string">"sample sizes don't add up"</span>.to_string(),
|
|
<a href=#121 id=121 data-nosnippet>121</a> ));
|
|
<a href=#122 id=122 data-nosnippet>122</a> }
|
|
<a href=#123 id=123 data-nosnippet>123</a>
|
|
<a href=#124 id=124 data-nosnippet>124</a> <span class="kw">let </span><span class="kw-2">mut </span>result = Vec::with_capacity(max_size);
|
|
<a href=#125 id=125 data-nosnippet>125</a> zstd_safe::train_from_buffer(<span class="kw-2">&mut </span>result, sample_data, sample_sizes)
|
|
<a href=#126 id=126 data-nosnippet>126</a> .map_err(map_error_code)<span class="question-mark">?</span>;
|
|
<a href=#127 id=127 data-nosnippet>127</a> <span class="prelude-val">Ok</span>(result)
|
|
<a href=#128 id=128 data-nosnippet>128</a>}
|
|
<a href=#129 id=129 data-nosnippet>129</a>
|
|
<a href=#130 id=130 data-nosnippet>130</a><span class="doccomment">/// Train a dictionary from multiple samples.
|
|
<a href=#131 id=131 data-nosnippet>131</a>///
|
|
<a href=#132 id=132 data-nosnippet>132</a>/// The samples will internally be copied to a single continuous buffer,
|
|
<a href=#133 id=133 data-nosnippet>133</a>/// so make sure you have enough memory available.
|
|
<a href=#134 id=134 data-nosnippet>134</a>///
|
|
<a href=#135 id=135 data-nosnippet>135</a>/// If you need to stretch your system's limits,
|
|
<a href=#136 id=136 data-nosnippet>136</a>/// [`from_continuous`] directly uses the given slice.
|
|
<a href=#137 id=137 data-nosnippet>137</a>///
|
|
<a href=#138 id=138 data-nosnippet>138</a>/// [`from_continuous`]: ./fn.from_continuous.html
|
|
<a href=#139 id=139 data-nosnippet>139</a>///
|
|
<a href=#140 id=140 data-nosnippet>140</a>/// * `samples` is a list of individual samples to train on.
|
|
<a href=#141 id=141 data-nosnippet>141</a>/// * `max_size` is the maximum size of the dictionary to generate.
|
|
<a href=#142 id=142 data-nosnippet>142</a>///
|
|
<a href=#143 id=143 data-nosnippet>143</a>/// The result is the dictionary data. You can, for example, feed it to [`CDict::create`].
|
|
<a href=#144 id=144 data-nosnippet>144</a></span><span class="attr">#[cfg(feature = <span class="string">"zdict_builder"</span>)]
|
|
<a href=#145 id=145 data-nosnippet>145</a>#[cfg_attr(feature = <span class="string">"doc-cfg"</span>, doc(cfg(feature = <span class="string">"zdict_builder"</span>)))]
|
|
<a href=#146 id=146 data-nosnippet>146</a></span><span class="kw">pub fn </span>from_samples<S: AsRef<[u8]>>(
|
|
<a href=#147 id=147 data-nosnippet>147</a> samples: <span class="kw-2">&</span>[S],
|
|
<a href=#148 id=148 data-nosnippet>148</a> max_size: usize,
|
|
<a href=#149 id=149 data-nosnippet>149</a>) -> io::Result<Vec<u8>> {
|
|
<a href=#150 id=150 data-nosnippet>150</a> <span class="comment">// Pre-allocate the entire required size.
|
|
<a href=#151 id=151 data-nosnippet>151</a> </span><span class="kw">let </span>total_length: usize =
|
|
<a href=#152 id=152 data-nosnippet>152</a> samples.iter().map(|sample| sample.as_ref().len()).sum();
|
|
<a href=#153 id=153 data-nosnippet>153</a>
|
|
<a href=#154 id=154 data-nosnippet>154</a> <span class="kw">let </span><span class="kw-2">mut </span>data = Vec::with_capacity(total_length);
|
|
<a href=#155 id=155 data-nosnippet>155</a>
|
|
<a href=#156 id=156 data-nosnippet>156</a> <span class="comment">// Copy every sample to a big chunk of memory
|
|
<a href=#157 id=157 data-nosnippet>157</a> </span>data.extend(samples.iter().flat_map(|s| s.as_ref()).cloned());
|
|
<a href=#158 id=158 data-nosnippet>158</a>
|
|
<a href=#159 id=159 data-nosnippet>159</a> <span class="kw">let </span>sizes: Vec<<span class="kw">_</span>> = samples.iter().map(|s| s.as_ref().len()).collect();
|
|
<a href=#160 id=160 data-nosnippet>160</a>
|
|
<a href=#161 id=161 data-nosnippet>161</a> from_continuous(<span class="kw-2">&</span>data, <span class="kw-2">&</span>sizes, max_size)
|
|
<a href=#162 id=162 data-nosnippet>162</a>}
|
|
<a href=#163 id=163 data-nosnippet>163</a>
|
|
<a href=#164 id=164 data-nosnippet>164</a><span class="doccomment">/// Train a dictionary from multiple samples.
|
|
<a href=#165 id=165 data-nosnippet>165</a>///
|
|
<a href=#166 id=166 data-nosnippet>166</a>/// Unlike [`from_samples`], this does not require having a list of all samples.
|
|
<a href=#167 id=167 data-nosnippet>167</a>/// It also allows running into an error when iterating through the samples.
|
|
<a href=#168 id=168 data-nosnippet>168</a>///
|
|
<a href=#169 id=169 data-nosnippet>169</a>/// They will still be copied to a continuous array and fed to [`from_continuous`].
|
|
<a href=#170 id=170 data-nosnippet>170</a>///
|
|
<a href=#171 id=171 data-nosnippet>171</a>/// * `samples` is an iterator of individual samples to train on.
|
|
<a href=#172 id=172 data-nosnippet>172</a>/// * `max_size` is the maximum size of the dictionary to generate.
|
|
<a href=#173 id=173 data-nosnippet>173</a>///
|
|
<a href=#174 id=174 data-nosnippet>174</a>/// The result is the dictionary data. You can, for example, feed it to [`CDict::create`].
|
|
<a href=#175 id=175 data-nosnippet>175</a>///
|
|
<a href=#176 id=176 data-nosnippet>176</a>/// # Examples
|
|
<a href=#177 id=177 data-nosnippet>177</a>///
|
|
<a href=#178 id=178 data-nosnippet>178</a>/// ```rust,no_run
|
|
<a href=#179 id=179 data-nosnippet>179</a>/// // Train from a couple of json files.
|
|
<a href=#180 id=180 data-nosnippet>180</a>/// let dict_buffer = zstd::dict::from_sample_iterator(
|
|
<a href=#181 id=181 data-nosnippet>181</a>/// ["file_a.json", "file_b.json"]
|
|
<a href=#182 id=182 data-nosnippet>182</a>/// .into_iter()
|
|
<a href=#183 id=183 data-nosnippet>183</a>/// .map(|filename| std::fs::File::open(filename)),
|
|
<a href=#184 id=184 data-nosnippet>184</a>/// 10_000, // 10kB dictionary
|
|
<a href=#185 id=185 data-nosnippet>185</a>/// ).unwrap();
|
|
<a href=#186 id=186 data-nosnippet>186</a>/// ```
|
|
<a href=#187 id=187 data-nosnippet>187</a>///
|
|
<a href=#188 id=188 data-nosnippet>188</a>/// ```rust,no_run
|
|
<a href=#189 id=189 data-nosnippet>189</a>/// use std::io::BufRead as _;
|
|
<a href=#190 id=190 data-nosnippet>190</a>/// // Treat each line from stdin as a separate sample.
|
|
<a href=#191 id=191 data-nosnippet>191</a>/// let dict_buffer = zstd::dict::from_sample_iterator(
|
|
<a href=#192 id=192 data-nosnippet>192</a>/// std::io::stdin().lock().lines().map(|line: std::io::Result<String>| {
|
|
<a href=#193 id=193 data-nosnippet>193</a>/// // Transform each line into a `Cursor<Vec<u8>>` so they implement Read.
|
|
<a href=#194 id=194 data-nosnippet>194</a>/// line.map(String::into_bytes)
|
|
<a href=#195 id=195 data-nosnippet>195</a>/// .map(std::io::Cursor::new)
|
|
<a href=#196 id=196 data-nosnippet>196</a>/// }),
|
|
<a href=#197 id=197 data-nosnippet>197</a>/// 10_000, // 10kB dictionary
|
|
<a href=#198 id=198 data-nosnippet>198</a>/// ).unwrap();
|
|
<a href=#199 id=199 data-nosnippet>199</a>/// ```
|
|
<a href=#200 id=200 data-nosnippet>200</a></span><span class="attr">#[cfg(feature = <span class="string">"zdict_builder"</span>)]
|
|
<a href=#201 id=201 data-nosnippet>201</a>#[cfg_attr(feature = <span class="string">"doc-cfg"</span>, doc(cfg(feature = <span class="string">"zdict_builder"</span>)))]
|
|
<a href=#202 id=202 data-nosnippet>202</a></span><span class="kw">pub fn </span>from_sample_iterator<I, R>(
|
|
<a href=#203 id=203 data-nosnippet>203</a> samples: I,
|
|
<a href=#204 id=204 data-nosnippet>204</a> max_size: usize,
|
|
<a href=#205 id=205 data-nosnippet>205</a>) -> io::Result<Vec<u8>>
|
|
<a href=#206 id=206 data-nosnippet>206</a><span class="kw">where
|
|
<a href=#207 id=207 data-nosnippet>207</a> </span>I: IntoIterator<Item = io::Result<R>>,
|
|
<a href=#208 id=208 data-nosnippet>208</a> R: Read,
|
|
<a href=#209 id=209 data-nosnippet>209</a>{
|
|
<a href=#210 id=210 data-nosnippet>210</a> <span class="kw">let </span><span class="kw-2">mut </span>data = Vec::new();
|
|
<a href=#211 id=211 data-nosnippet>211</a> <span class="kw">let </span><span class="kw-2">mut </span>sizes = Vec::new();
|
|
<a href=#212 id=212 data-nosnippet>212</a>
|
|
<a href=#213 id=213 data-nosnippet>213</a> <span class="kw">for </span>sample <span class="kw">in </span>samples {
|
|
<a href=#214 id=214 data-nosnippet>214</a> <span class="kw">let </span><span class="kw-2">mut </span>sample = sample<span class="question-mark">?</span>;
|
|
<a href=#215 id=215 data-nosnippet>215</a> <span class="kw">let </span>len = sample.read_to_end(<span class="kw-2">&mut </span>data)<span class="question-mark">?</span>;
|
|
<a href=#216 id=216 data-nosnippet>216</a> sizes.push(len);
|
|
<a href=#217 id=217 data-nosnippet>217</a> }
|
|
<a href=#218 id=218 data-nosnippet>218</a>
|
|
<a href=#219 id=219 data-nosnippet>219</a> from_continuous(<span class="kw-2">&</span>data, <span class="kw-2">&</span>sizes, max_size)
|
|
<a href=#220 id=220 data-nosnippet>220</a>}
|
|
<a href=#221 id=221 data-nosnippet>221</a>
|
|
<a href=#222 id=222 data-nosnippet>222</a><span class="doccomment">/// Train a dict from a list of files.
|
|
<a href=#223 id=223 data-nosnippet>223</a>///
|
|
<a href=#224 id=224 data-nosnippet>224</a>/// * `filenames` is an iterator of files to load. Each file will be treated as an individual
|
|
<a href=#225 id=225 data-nosnippet>225</a>/// sample.
|
|
<a href=#226 id=226 data-nosnippet>226</a>/// * `max_size` is the maximum size of the dictionary to generate.
|
|
<a href=#227 id=227 data-nosnippet>227</a>///
|
|
<a href=#228 id=228 data-nosnippet>228</a>/// The result is the dictionary data. You can, for example, feed it to [`CDict::create`].
|
|
<a href=#229 id=229 data-nosnippet>229</a></span><span class="attr">#[cfg(feature = <span class="string">"zdict_builder"</span>)]
|
|
<a href=#230 id=230 data-nosnippet>230</a>#[cfg_attr(feature = <span class="string">"doc-cfg"</span>, doc(cfg(feature = <span class="string">"zdict_builder"</span>)))]
|
|
<a href=#231 id=231 data-nosnippet>231</a></span><span class="kw">pub fn </span>from_files<I, P>(filenames: I, max_size: usize) -> io::Result<Vec<u8>>
|
|
<a href=#232 id=232 data-nosnippet>232</a><span class="kw">where
|
|
<a href=#233 id=233 data-nosnippet>233</a> </span>P: AsRef<std::path::Path>,
|
|
<a href=#234 id=234 data-nosnippet>234</a> I: IntoIterator<Item = P>,
|
|
<a href=#235 id=235 data-nosnippet>235</a>{
|
|
<a href=#236 id=236 data-nosnippet>236</a> from_sample_iterator(
|
|
<a href=#237 id=237 data-nosnippet>237</a> filenames
|
|
<a href=#238 id=238 data-nosnippet>238</a> .into_iter()
|
|
<a href=#239 id=239 data-nosnippet>239</a> .map(|filename| std::fs::File::open(filename)),
|
|
<a href=#240 id=240 data-nosnippet>240</a> max_size,
|
|
<a href=#241 id=241 data-nosnippet>241</a> )
|
|
<a href=#242 id=242 data-nosnippet>242</a>}
|
|
<a href=#243 id=243 data-nosnippet>243</a>
|
|
<a href=#244 id=244 data-nosnippet>244</a><span class="attr">#[cfg(test)]
|
|
<a href=#245 id=245 data-nosnippet>245</a>#[cfg(feature = <span class="string">"zdict_builder"</span>)]
|
|
<a href=#246 id=246 data-nosnippet>246</a></span><span class="kw">mod </span>tests {
|
|
<a href=#247 id=247 data-nosnippet>247</a> <span class="kw">use </span>std::fs;
|
|
<a href=#248 id=248 data-nosnippet>248</a> <span class="kw">use </span>std::io;
|
|
<a href=#249 id=249 data-nosnippet>249</a> <span class="kw">use </span>std::io::Read;
|
|
<a href=#250 id=250 data-nosnippet>250</a>
|
|
<a href=#251 id=251 data-nosnippet>251</a> <span class="kw">use </span>walkdir;
|
|
<a href=#252 id=252 data-nosnippet>252</a>
|
|
<a href=#253 id=253 data-nosnippet>253</a> <span class="attr">#[test]
|
|
<a href=#254 id=254 data-nosnippet>254</a> </span><span class="kw">fn </span>test_dict_training() {
|
|
<a href=#255 id=255 data-nosnippet>255</a> <span class="comment">// Train a dictionary
|
|
<a href=#256 id=256 data-nosnippet>256</a> </span><span class="kw">let </span>paths: Vec<<span class="kw">_</span>> = walkdir::WalkDir::new(<span class="string">"src"</span>)
|
|
<a href=#257 id=257 data-nosnippet>257</a> .into_iter()
|
|
<a href=#258 id=258 data-nosnippet>258</a> .map(|entry| entry.unwrap())
|
|
<a href=#259 id=259 data-nosnippet>259</a> .map(|entry| entry.into_path())
|
|
<a href=#260 id=260 data-nosnippet>260</a> .filter(|path| path.to_str().unwrap().ends_with(<span class="string">".rs"</span>))
|
|
<a href=#261 id=261 data-nosnippet>261</a> .collect();
|
|
<a href=#262 id=262 data-nosnippet>262</a>
|
|
<a href=#263 id=263 data-nosnippet>263</a> <span class="kw">let </span>dict = <span class="kw">super</span>::from_files(<span class="kw-2">&</span>paths, <span class="number">4000</span>).unwrap();
|
|
<a href=#264 id=264 data-nosnippet>264</a>
|
|
<a href=#265 id=265 data-nosnippet>265</a> <span class="kw">for </span>path <span class="kw">in </span>paths {
|
|
<a href=#266 id=266 data-nosnippet>266</a> <span class="kw">let </span><span class="kw-2">mut </span>buffer = Vec::new();
|
|
<a href=#267 id=267 data-nosnippet>267</a> <span class="kw">let </span><span class="kw-2">mut </span>file = fs::File::open(path).unwrap();
|
|
<a href=#268 id=268 data-nosnippet>268</a> <span class="kw">let </span><span class="kw-2">mut </span>content = Vec::new();
|
|
<a href=#269 id=269 data-nosnippet>269</a> file.read_to_end(<span class="kw-2">&mut </span>content).unwrap();
|
|
<a href=#270 id=270 data-nosnippet>270</a> io::copy(
|
|
<a href=#271 id=271 data-nosnippet>271</a> <span class="kw-2">&mut &</span>content[..],
|
|
<a href=#272 id=272 data-nosnippet>272</a> <span class="kw-2">&mut </span><span class="kw">crate</span>::stream::Encoder::with_dictionary(
|
|
<a href=#273 id=273 data-nosnippet>273</a> <span class="kw-2">&mut </span>buffer,
|
|
<a href=#274 id=274 data-nosnippet>274</a> <span class="number">1</span>,
|
|
<a href=#275 id=275 data-nosnippet>275</a> <span class="kw-2">&</span>dict,
|
|
<a href=#276 id=276 data-nosnippet>276</a> )
|
|
<a href=#277 id=277 data-nosnippet>277</a> .unwrap()
|
|
<a href=#278 id=278 data-nosnippet>278</a> .auto_finish(),
|
|
<a href=#279 id=279 data-nosnippet>279</a> )
|
|
<a href=#280 id=280 data-nosnippet>280</a> .unwrap();
|
|
<a href=#281 id=281 data-nosnippet>281</a>
|
|
<a href=#282 id=282 data-nosnippet>282</a> <span class="kw">let </span><span class="kw-2">mut </span>result = Vec::new();
|
|
<a href=#283 id=283 data-nosnippet>283</a> io::copy(
|
|
<a href=#284 id=284 data-nosnippet>284</a> <span class="kw-2">&mut </span><span class="kw">crate</span>::stream::Decoder::with_dictionary(
|
|
<a href=#285 id=285 data-nosnippet>285</a> <span class="kw-2">&</span>buffer[..],
|
|
<a href=#286 id=286 data-nosnippet>286</a> <span class="kw-2">&</span>dict[..],
|
|
<a href=#287 id=287 data-nosnippet>287</a> )
|
|
<a href=#288 id=288 data-nosnippet>288</a> .unwrap(),
|
|
<a href=#289 id=289 data-nosnippet>289</a> <span class="kw-2">&mut </span>result,
|
|
<a href=#290 id=290 data-nosnippet>290</a> )
|
|
<a href=#291 id=291 data-nosnippet>291</a> .unwrap();
|
|
<a href=#292 id=292 data-nosnippet>292</a>
|
|
<a href=#293 id=293 data-nosnippet>293</a> <span class="macro">assert_eq!</span>(<span class="kw-2">&</span>content, <span class="kw-2">&</span>result);
|
|
<a href=#294 id=294 data-nosnippet>294</a> }
|
|
<a href=#295 id=295 data-nosnippet>295</a> }
|
|
<a href=#296 id=296 data-nosnippet>296</a>}
|
|
</code></pre></div></section></main></body></html> |