Added: dev/commons/csv/1.7-RC1/site/apidocs/src-html/org/apache/commons/csv/CSVParser.html ============================================================================== --- dev/commons/csv/1.7-RC1/site/apidocs/src-html/org/apache/commons/csv/CSVParser.html (added) +++ dev/commons/csv/1.7-RC1/site/apidocs/src-html/org/apache/commons/csv/CSVParser.html Sun Jun 2 00:19:07 2019 @@ -0,0 +1,774 @@ +<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd";> +<html lang="en"> +<head> +<title>Source code</title> +<link rel="stylesheet" type="text/css" href="../../../../../stylesheet.css" title="Style"> +</head> +<body> +<div class="sourceContainer"> +<pre><span class="sourceLineNo">001</span>/*<a name="line.1"></a> +<span class="sourceLineNo">002</span> * Licensed to the Apache Software Foundation (ASF) under one or more<a name="line.2"></a> +<span class="sourceLineNo">003</span> * contributor license agreements. See the NOTICE file distributed with<a name="line.3"></a> +<span class="sourceLineNo">004</span> * this work for additional information regarding copyright ownership.<a name="line.4"></a> +<span class="sourceLineNo">005</span> * The ASF licenses this file to You under the Apache License, Version 2.0<a name="line.5"></a> +<span class="sourceLineNo">006</span> * (the "License"); you may not use this file except in compliance with<a name="line.6"></a> +<span class="sourceLineNo">007</span> * the License. You may obtain a copy of the License at<a name="line.7"></a> +<span class="sourceLineNo">008</span> *<a name="line.8"></a> +<span class="sourceLineNo">009</span> * http://www.apache.org/licenses/LICENSE-2.0<a name="line.9"></a> +<span class="sourceLineNo">010</span> *<a name="line.10"></a> +<span class="sourceLineNo">011</span> * Unless required by applicable law or agreed to in writing, software<a name="line.11"></a> +<span class="sourceLineNo">012</span> * distributed under the License is distributed on an "AS IS" BASIS,<a name="line.12"></a> +<span class="sourceLineNo">013</span> * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.<a name="line.13"></a> +<span class="sourceLineNo">014</span> * See the License for the specific language governing permissions and<a name="line.14"></a> +<span class="sourceLineNo">015</span> * limitations under the License.<a name="line.15"></a> +<span class="sourceLineNo">016</span> */<a name="line.16"></a> +<span class="sourceLineNo">017</span><a name="line.17"></a> +<span class="sourceLineNo">018</span>package org.apache.commons.csv;<a name="line.18"></a> +<span class="sourceLineNo">019</span><a name="line.19"></a> +<span class="sourceLineNo">020</span>import static org.apache.commons.csv.Token.Type.TOKEN;<a name="line.20"></a> +<span class="sourceLineNo">021</span><a name="line.21"></a> +<span class="sourceLineNo">022</span>import java.io.Closeable;<a name="line.22"></a> +<span class="sourceLineNo">023</span>import java.io.File;<a name="line.23"></a> +<span class="sourceLineNo">024</span>import java.io.FileInputStream;<a name="line.24"></a> +<span class="sourceLineNo">025</span>import java.io.IOException;<a name="line.25"></a> +<span class="sourceLineNo">026</span>import java.io.InputStream;<a name="line.26"></a> +<span class="sourceLineNo">027</span>import java.io.InputStreamReader;<a name="line.27"></a> +<span class="sourceLineNo">028</span>import java.io.Reader;<a name="line.28"></a> +<span class="sourceLineNo">029</span>import java.io.StringReader;<a name="line.29"></a> +<span class="sourceLineNo">030</span>import java.net.URL;<a name="line.30"></a> +<span class="sourceLineNo">031</span>import java.nio.charset.Charset;<a name="line.31"></a> +<span class="sourceLineNo">032</span>import java.nio.file.Files;<a name="line.32"></a> +<span class="sourceLineNo">033</span>import java.nio.file.Path;<a name="line.33"></a> +<span class="sourceLineNo">034</span>import java.util.ArrayList;<a name="line.34"></a> +<span class="sourceLineNo">035</span>import java.util.Arrays;<a name="line.35"></a> +<span class="sourceLineNo">036</span>import java.util.Collections;<a name="line.36"></a> +<span class="sourceLineNo">037</span>import java.util.Iterator;<a name="line.37"></a> +<span class="sourceLineNo">038</span>import java.util.LinkedHashMap;<a name="line.38"></a> +<span class="sourceLineNo">039</span>import java.util.List;<a name="line.39"></a> +<span class="sourceLineNo">040</span>import java.util.Map;<a name="line.40"></a> +<span class="sourceLineNo">041</span>import java.util.NoSuchElementException;<a name="line.41"></a> +<span class="sourceLineNo">042</span>import java.util.TreeMap;<a name="line.42"></a> +<span class="sourceLineNo">043</span><a name="line.43"></a> +<span class="sourceLineNo">044</span>/**<a name="line.44"></a> +<span class="sourceLineNo">045</span> * Parses CSV files according to the specified format.<a name="line.45"></a> +<span class="sourceLineNo">046</span> *<a name="line.46"></a> +<span class="sourceLineNo">047</span> * Because CSV appears in many different dialects, the parser supports many formats by allowing the<a name="line.47"></a> +<span class="sourceLineNo">048</span> * specification of a {@link CSVFormat}.<a name="line.48"></a> +<span class="sourceLineNo">049</span> *<a name="line.49"></a> +<span class="sourceLineNo">050</span> * The parser works record wise. It is not possible to go back, once a record has been parsed from the input stream.<a name="line.50"></a> +<span class="sourceLineNo">051</span> *<a name="line.51"></a> +<span class="sourceLineNo">052</span> * <h2>Creating instances</h2><a name="line.52"></a> +<span class="sourceLineNo">053</span> * <p><a name="line.53"></a> +<span class="sourceLineNo">054</span> * There are several static factory methods that can be used to create instances for various types of resources:<a name="line.54"></a> +<span class="sourceLineNo">055</span> * </p><a name="line.55"></a> +<span class="sourceLineNo">056</span> * <ul><a name="line.56"></a> +<span class="sourceLineNo">057</span> * <li>{@link #parse(java.io.File, Charset, CSVFormat)}</li><a name="line.57"></a> +<span class="sourceLineNo">058</span> * <li>{@link #parse(String, CSVFormat)}</li><a name="line.58"></a> +<span class="sourceLineNo">059</span> * <li>{@link #parse(java.net.URL, java.nio.charset.Charset, CSVFormat)}</li><a name="line.59"></a> +<span class="sourceLineNo">060</span> * </ul><a name="line.60"></a> +<span class="sourceLineNo">061</span> * <p><a name="line.61"></a> +<span class="sourceLineNo">062</span> * Alternatively parsers can also be created by passing a {@link Reader} directly to the sole constructor.<a name="line.62"></a> +<span class="sourceLineNo">063</span> *<a name="line.63"></a> +<span class="sourceLineNo">064</span> * For those who like fluent APIs, parsers can be created using {@link CSVFormat#parse(java.io.Reader)} as a shortcut:<a name="line.64"></a> +<span class="sourceLineNo">065</span> * </p><a name="line.65"></a> +<span class="sourceLineNo">066</span> * <pre><a name="line.66"></a> +<span class="sourceLineNo">067</span> * for(CSVRecord record : CSVFormat.EXCEL.parse(in)) {<a name="line.67"></a> +<span class="sourceLineNo">068</span> * ...<a name="line.68"></a> +<span class="sourceLineNo">069</span> * }<a name="line.69"></a> +<span class="sourceLineNo">070</span> * </pre><a name="line.70"></a> +<span class="sourceLineNo">071</span> *<a name="line.71"></a> +<span class="sourceLineNo">072</span> * <h2>Parsing record wise</h2><a name="line.72"></a> +<span class="sourceLineNo">073</span> * <p><a name="line.73"></a> +<span class="sourceLineNo">074</span> * To parse a CSV input from a file, you write:<a name="line.74"></a> +<span class="sourceLineNo">075</span> * </p><a name="line.75"></a> +<span class="sourceLineNo">076</span> *<a name="line.76"></a> +<span class="sourceLineNo">077</span> * <pre><a name="line.77"></a> +<span class="sourceLineNo">078</span> * File csvData = new File(&quot;/path/to/csv&quot;);<a name="line.78"></a> +<span class="sourceLineNo">079</span> * CSVParser parser = CSVParser.parse(csvData, CSVFormat.RFC4180);<a name="line.79"></a> +<span class="sourceLineNo">080</span> * for (CSVRecord csvRecord : parser) {<a name="line.80"></a> +<span class="sourceLineNo">081</span> * ...<a name="line.81"></a> +<span class="sourceLineNo">082</span> * }<a name="line.82"></a> +<span class="sourceLineNo">083</span> * </pre><a name="line.83"></a> +<span class="sourceLineNo">084</span> *<a name="line.84"></a> +<span class="sourceLineNo">085</span> * <p><a name="line.85"></a> +<span class="sourceLineNo">086</span> * This will read the parse the contents of the file using the<a name="line.86"></a> +<span class="sourceLineNo">087</span> * <a href="http://tools.ietf.org/html/rfc4180"; target="_blank">RFC 4180</a> format.<a name="line.87"></a> +<span class="sourceLineNo">088</span> * </p><a name="line.88"></a> +<span class="sourceLineNo">089</span> *<a name="line.89"></a> +<span class="sourceLineNo">090</span> * <p><a name="line.90"></a> +<span class="sourceLineNo">091</span> * To parse CSV input in a format like Excel, you write:<a name="line.91"></a> +<span class="sourceLineNo">092</span> * </p><a name="line.92"></a> +<span class="sourceLineNo">093</span> *<a name="line.93"></a> +<span class="sourceLineNo">094</span> * <pre><a name="line.94"></a> +<span class="sourceLineNo">095</span> * CSVParser parser = CSVParser.parse(csvData, CSVFormat.EXCEL);<a name="line.95"></a> +<span class="sourceLineNo">096</span> * for (CSVRecord csvRecord : parser) {<a name="line.96"></a> +<span class="sourceLineNo">097</span> * ...<a name="line.97"></a> +<span class="sourceLineNo">098</span> * }<a name="line.98"></a> +<span class="sourceLineNo">099</span> * </pre><a name="line.99"></a> +<span class="sourceLineNo">100</span> *<a name="line.100"></a> +<span class="sourceLineNo">101</span> * <p><a name="line.101"></a> +<span class="sourceLineNo">102</span> * If the predefined formats don't match the format at hands, custom formats can be defined. More information about<a name="line.102"></a> +<span class="sourceLineNo">103</span> * customising CSVFormats is available in {@link CSVFormat CSVFormat Javadoc}.<a name="line.103"></a> +<span class="sourceLineNo">104</span> * </p><a name="line.104"></a> +<span class="sourceLineNo">105</span> *<a name="line.105"></a> +<span class="sourceLineNo">106</span> * <h2>Parsing into memory</h2><a name="line.106"></a> +<span class="sourceLineNo">107</span> * <p><a name="line.107"></a> +<span class="sourceLineNo">108</span> * If parsing record wise is not desired, the contents of the input can be read completely into memory.<a name="line.108"></a> +<span class="sourceLineNo">109</span> * </p><a name="line.109"></a> +<span class="sourceLineNo">110</span> *<a name="line.110"></a> +<span class="sourceLineNo">111</span> * <pre><a name="line.111"></a> +<span class="sourceLineNo">112</span> * Reader in = new StringReader(&quot;a;b\nc;d&quot;);<a name="line.112"></a> +<span class="sourceLineNo">113</span> * CSVParser parser = new CSVParser(in, CSVFormat.EXCEL);<a name="line.113"></a> +<span class="sourceLineNo">114</span> * List&lt;CSVRecord&gt; list = parser.getRecords();<a name="line.114"></a> +<span class="sourceLineNo">115</span> * </pre><a name="line.115"></a> +<span class="sourceLineNo">116</span> *<a name="line.116"></a> +<span class="sourceLineNo">117</span> * <p><a name="line.117"></a> +<span class="sourceLineNo">118</span> * There are two constraints that have to be kept in mind:<a name="line.118"></a> +<span class="sourceLineNo">119</span> * </p><a name="line.119"></a> +<span class="sourceLineNo">120</span> *<a name="line.120"></a> +<span class="sourceLineNo">121</span> * <ol><a name="line.121"></a> +<span class="sourceLineNo">122</span> * <li>Parsing into memory starts at the current position of the parser. If you have already parsed records from<a name="line.122"></a> +<span class="sourceLineNo">123</span> * the input, those records will not end up in the in memory representation of your CSV data.</li><a name="line.123"></a> +<span class="sourceLineNo">124</span> * <li>Parsing into memory may consume a lot of system resources depending on the input. For example if you're<a name="line.124"></a> +<span class="sourceLineNo">125</span> * parsing a 150MB file of CSV data the contents will be read completely into memory.</li><a name="line.125"></a> +<span class="sourceLineNo">126</span> * </ol><a name="line.126"></a> +<span class="sourceLineNo">127</span> *<a name="line.127"></a> +<span class="sourceLineNo">128</span> * <h2>Notes</h2><a name="line.128"></a> +<span class="sourceLineNo">129</span> * <p><a name="line.129"></a> +<span class="sourceLineNo">130</span> * Internal parser state is completely covered by the format and the reader-state.<a name="line.130"></a> +<span class="sourceLineNo">131</span> * </p><a name="line.131"></a> +<span class="sourceLineNo">132</span> *<a name="line.132"></a> +<span class="sourceLineNo">133</span> * @see <a href="package-summary.html">package documentation for more details</a><a name="line.133"></a> +<span class="sourceLineNo">134</span> */<a name="line.134"></a> +<span class="sourceLineNo">135</span>public final class CSVParser implements Iterable<CSVRecord>, Closeable {<a name="line.135"></a> +<span class="sourceLineNo">136</span><a name="line.136"></a> +<span class="sourceLineNo">137</span> class CSVRecordIterator implements Iterator<CSVRecord> {<a name="line.137"></a> +<span class="sourceLineNo">138</span> private CSVRecord current;<a name="line.138"></a> +<span class="sourceLineNo">139</span><a name="line.139"></a> +<span class="sourceLineNo">140</span> private CSVRecord getNextRecord() {<a name="line.140"></a> +<span class="sourceLineNo">141</span> try {<a name="line.141"></a> +<span class="sourceLineNo">142</span> return CSVParser.this.nextRecord();<a name="line.142"></a> +<span class="sourceLineNo">143</span> } catch (final IOException e) {<a name="line.143"></a> +<span class="sourceLineNo">144</span> throw new IllegalStateException(<a name="line.144"></a> +<span class="sourceLineNo">145</span> e.getClass().getSimpleName() + " reading next record: " + e.toString(), e);<a name="line.145"></a> +<span class="sourceLineNo">146</span> }<a name="line.146"></a> +<span class="sourceLineNo">147</span> }<a name="line.147"></a> +<span class="sourceLineNo">148</span><a name="line.148"></a> +<span class="sourceLineNo">149</span> @Override<a name="line.149"></a> +<span class="sourceLineNo">150</span> public boolean hasNext() {<a name="line.150"></a> +<span class="sourceLineNo">151</span> if (CSVParser.this.isClosed()) {<a name="line.151"></a> +<span class="sourceLineNo">152</span> return false;<a name="line.152"></a> +<span class="sourceLineNo">153</span> }<a name="line.153"></a> +<span class="sourceLineNo">154</span> if (this.current == null) {<a name="line.154"></a> +<span class="sourceLineNo">155</span> this.current = this.getNextRecord();<a name="line.155"></a> +<span class="sourceLineNo">156</span> }<a name="line.156"></a> +<span class="sourceLineNo">157</span><a name="line.157"></a> +<span class="sourceLineNo">158</span> return this.current != null;<a name="line.158"></a> +<span class="sourceLineNo">159</span> }<a name="line.159"></a> +<span class="sourceLineNo">160</span><a name="line.160"></a> +<span class="sourceLineNo">161</span> @Override<a name="line.161"></a> +<span class="sourceLineNo">162</span> public CSVRecord next() {<a name="line.162"></a> +<span class="sourceLineNo">163</span> if (CSVParser.this.isClosed()) {<a name="line.163"></a> +<span class="sourceLineNo">164</span> throw new NoSuchElementException("CSVParser has been closed");<a name="line.164"></a> +<span class="sourceLineNo">165</span> }<a name="line.165"></a> +<span class="sourceLineNo">166</span> CSVRecord next = this.current;<a name="line.166"></a> +<span class="sourceLineNo">167</span> this.current = null;<a name="line.167"></a> +<span class="sourceLineNo">168</span><a name="line.168"></a> +<span class="sourceLineNo">169</span> if (next == null) {<a name="line.169"></a> +<span class="sourceLineNo">170</span> // hasNext() wasn't called before<a name="line.170"></a> +<span class="sourceLineNo">171</span> next = this.getNextRecord();<a name="line.171"></a> +<span class="sourceLineNo">172</span> if (next == null) {<a name="line.172"></a> +<span class="sourceLineNo">173</span> throw new NoSuchElementException("No more CSV records available");<a name="line.173"></a> +<span class="sourceLineNo">174</span> }<a name="line.174"></a> +<span class="sourceLineNo">175</span> }<a name="line.175"></a> +<span class="sourceLineNo">176</span><a name="line.176"></a> +<span class="sourceLineNo">177</span> return next;<a name="line.177"></a> +<span class="sourceLineNo">178</span> }<a name="line.178"></a> +<span class="sourceLineNo">179</span><a name="line.179"></a> +<span class="sourceLineNo">180</span> @Override<a name="line.180"></a> +<span class="sourceLineNo">181</span> public void remove() {<a name="line.181"></a> +<span class="sourceLineNo">182</span> throw new UnsupportedOperationException();<a name="line.182"></a> +<span class="sourceLineNo">183</span> }<a name="line.183"></a> +<span class="sourceLineNo">184</span> }<a name="line.184"></a> +<span class="sourceLineNo">185</span><a name="line.185"></a> +<span class="sourceLineNo">186</span> /**<a name="line.186"></a> +<span class="sourceLineNo">187</span> * Creates a parser for the given {@link File}.<a name="line.187"></a> +<span class="sourceLineNo">188</span> *<a name="line.188"></a> +<span class="sourceLineNo">189</span> * @param file<a name="line.189"></a> +<span class="sourceLineNo">190</span> * a CSV file. Must not be null.<a name="line.190"></a> +<span class="sourceLineNo">191</span> * @param charset<a name="line.191"></a> +<span class="sourceLineNo">192</span> * A Charset<a name="line.192"></a> +<span class="sourceLineNo">193</span> * @param format<a name="line.193"></a> +<span class="sourceLineNo">194</span> * the CSVFormat used for CSV parsing. Must not be null.<a name="line.194"></a> +<span class="sourceLineNo">195</span> * @return a new parser<a name="line.195"></a> +<span class="sourceLineNo">196</span> * @throws IllegalArgumentException<a name="line.196"></a> +<span class="sourceLineNo">197</span> * If the parameters of the format are inconsistent or if either file or format are null.<a name="line.197"></a> +<span class="sourceLineNo">198</span> * @throws IOException<a name="line.198"></a> +<span class="sourceLineNo">199</span> * If an I/O error occurs<a name="line.199"></a> +<span class="sourceLineNo">200</span> */<a name="line.200"></a> +<span class="sourceLineNo">201</span> @SuppressWarnings("resource")<a name="line.201"></a> +<span class="sourceLineNo">202</span> public static CSVParser parse(final File file, final Charset charset, final CSVFormat format) throws IOException {<a name="line.202"></a> +<span class="sourceLineNo">203</span> Assertions.notNull(file, "file");<a name="line.203"></a> +<span class="sourceLineNo">204</span> Assertions.notNull(format, "format");<a name="line.204"></a> +<span class="sourceLineNo">205</span> return new CSVParser(new InputStreamReader(new FileInputStream(file), charset), format);<a name="line.205"></a> +<span class="sourceLineNo">206</span> }<a name="line.206"></a> +<span class="sourceLineNo">207</span><a name="line.207"></a> +<span class="sourceLineNo">208</span> /**<a name="line.208"></a> +<span class="sourceLineNo">209</span> * Creates a CSV parser using the given {@link CSVFormat}.<a name="line.209"></a> +<span class="sourceLineNo">210</span> *<a name="line.210"></a> +<span class="sourceLineNo">211</span> * <p><a name="line.211"></a> +<span class="sourceLineNo">212</span> * If you do not read all records from the given {@code reader}, you should call {@link #close()} on the parser,<a name="line.212"></a> +<span class="sourceLineNo">213</span> * unless you close the {@code reader}.<a name="line.213"></a> +<span class="sourceLineNo">214</span> * </p><a name="line.214"></a> +<span class="sourceLineNo">215</span> *<a name="line.215"></a> +<span class="sourceLineNo">216</span> * @param inputStream<a name="line.216"></a> +<span class="sourceLineNo">217</span> * an InputStream containing CSV-formatted input. Must not be null.<a name="line.217"></a> +<span class="sourceLineNo">218</span> * @param charset<a name="line.218"></a> +<span class="sourceLineNo">219</span> * a Charset.<a name="line.219"></a> +<span class="sourceLineNo">220</span> * @param format<a name="line.220"></a> +<span class="sourceLineNo">221</span> * the CSVFormat used for CSV parsing. Must not be null.<a name="line.221"></a> +<span class="sourceLineNo">222</span> * @return a new CSVParser configured with the given reader and format.<a name="line.222"></a> +<span class="sourceLineNo">223</span> * @throws IllegalArgumentException<a name="line.223"></a> +<span class="sourceLineNo">224</span> * If the parameters of the format are inconsistent or if either reader or format are null.<a name="line.224"></a> +<span class="sourceLineNo">225</span> * @throws IOException<a name="line.225"></a> +<span class="sourceLineNo">226</span> * If there is a problem reading the header or skipping the first record<a name="line.226"></a> +<span class="sourceLineNo">227</span> * @since 1.5<a name="line.227"></a> +<span class="sourceLineNo">228</span> */<a name="line.228"></a> +<span class="sourceLineNo">229</span> @SuppressWarnings("resource")<a name="line.229"></a> +<span class="sourceLineNo">230</span> public static CSVParser parse(final InputStream inputStream, final Charset charset, final CSVFormat format)<a name="line.230"></a> +<span class="sourceLineNo">231</span> throws IOException {<a name="line.231"></a> +<span class="sourceLineNo">232</span> Assertions.notNull(inputStream, "inputStream");<a name="line.232"></a> +<span class="sourceLineNo">233</span> Assertions.notNull(format, "format");<a name="line.233"></a> +<span class="sourceLineNo">234</span> return parse(new InputStreamReader(inputStream, charset), format);<a name="line.234"></a> +<span class="sourceLineNo">235</span> }<a name="line.235"></a> +<span class="sourceLineNo">236</span><a name="line.236"></a> +<span class="sourceLineNo">237</span> /**<a name="line.237"></a> +<span class="sourceLineNo">238</span> * Creates a parser for the given {@link Path}.<a name="line.238"></a> +<span class="sourceLineNo">239</span> *<a name="line.239"></a> +<span class="sourceLineNo">240</span> * @param path<a name="line.240"></a> +<span class="sourceLineNo">241</span> * a CSV file. Must not be null.<a name="line.241"></a> +<span class="sourceLineNo">242</span> * @param charset<a name="line.242"></a> +<span class="sourceLineNo">243</span> * A Charset<a name="line.243"></a> +<span class="sourceLineNo">244</span> * @param format<a name="line.244"></a> +<span class="sourceLineNo">245</span> * the CSVFormat used for CSV parsing. Must not be null.<a name="line.245"></a> +<span class="sourceLineNo">246</span> * @return a new parser<a name="line.246"></a> +<span class="sourceLineNo">247</span> * @throws IllegalArgumentException<a name="line.247"></a> +<span class="sourceLineNo">248</span> * If the parameters of the format are inconsistent or if either file or format are null.<a name="line.248"></a> +<span class="sourceLineNo">249</span> * @throws IOException<a name="line.249"></a> +<span class="sourceLineNo">250</span> * If an I/O error occurs<a name="line.250"></a> +<span class="sourceLineNo">251</span> * @since 1.5<a name="line.251"></a> +<span class="sourceLineNo">252</span> */<a name="line.252"></a> +<span class="sourceLineNo">253</span> public static CSVParser parse(final Path path, final Charset charset, final CSVFormat format) throws IOException {<a name="line.253"></a> +<span class="sourceLineNo">254</span> Assertions.notNull(path, "path");<a name="line.254"></a> +<span class="sourceLineNo">255</span> Assertions.notNull(format, "format");<a name="line.255"></a> +<span class="sourceLineNo">256</span> return parse(Files.newInputStream(path), charset, format);<a name="line.256"></a> +<span class="sourceLineNo">257</span> }<a name="line.257"></a> +<span class="sourceLineNo">258</span><a name="line.258"></a> +<span class="sourceLineNo">259</span> /**<a name="line.259"></a> +<span class="sourceLineNo">260</span> * Creates a CSV parser using the given {@link CSVFormat}<a name="line.260"></a> +<span class="sourceLineNo">261</span> *<a name="line.261"></a> +<span class="sourceLineNo">262</span> * <p><a name="line.262"></a> +<span class="sourceLineNo">263</span> * If you do not read all records from the given {@code reader}, you should call {@link #close()} on the parser,<a name="line.263"></a> +<span class="sourceLineNo">264</span> * unless you close the {@code reader}.<a name="line.264"></a> +<span class="sourceLineNo">265</span> * </p><a name="line.265"></a> +<span class="sourceLineNo">266</span> *<a name="line.266"></a> +<span class="sourceLineNo">267</span> * @param reader<a name="line.267"></a> +<span class="sourceLineNo">268</span> * a Reader containing CSV-formatted input. Must not be null.<a name="line.268"></a> +<span class="sourceLineNo">269</span> * @param format<a name="line.269"></a> +<span class="sourceLineNo">270</span> * the CSVFormat used for CSV parsing. Must not be null.<a name="line.270"></a> +<span class="sourceLineNo">271</span> * @return a new CSVParser configured with the given reader and format.<a name="line.271"></a> +<span class="sourceLineNo">272</span> * @throws IllegalArgumentException<a name="line.272"></a> +<span class="sourceLineNo">273</span> * If the parameters of the format are inconsistent or if either reader or format are null.<a name="line.273"></a> +<span class="sourceLineNo">274</span> * @throws IOException<a name="line.274"></a> +<span class="sourceLineNo">275</span> * If there is a problem reading the header or skipping the first record<a name="line.275"></a> +<span class="sourceLineNo">276</span> * @since 1.5<a name="line.276"></a> +<span class="sourceLineNo">277</span> */<a name="line.277"></a> +<span class="sourceLineNo">278</span> public static CSVParser parse(final Reader reader, final CSVFormat format) throws IOException {<a name="line.278"></a> +<span class="sourceLineNo">279</span> return new CSVParser(reader, format);<a name="line.279"></a> +<span class="sourceLineNo">280</span> }<a name="line.280"></a> +<span class="sourceLineNo">281</span><a name="line.281"></a> +<span class="sourceLineNo">282</span> /**<a name="line.282"></a> +<span class="sourceLineNo">283</span> * Creates a parser for the given {@link String}.<a name="line.283"></a> +<span class="sourceLineNo">284</span> *<a name="line.284"></a> +<span class="sourceLineNo">285</span> * @param string<a name="line.285"></a> +<span class="sourceLineNo">286</span> * a CSV string. Must not be null.<a name="line.286"></a> +<span class="sourceLineNo">287</span> * @param format<a name="line.287"></a> +<span class="sourceLineNo">288</span> * the CSVFormat used for CSV parsing. Must not be null.<a name="line.288"></a> +<span class="sourceLineNo">289</span> * @return a new parser<a name="line.289"></a> +<span class="sourceLineNo">290</span> * @throws IllegalArgumentException<a name="line.290"></a> +<span class="sourceLineNo">291</span> * If the parameters of the format are inconsistent or if either string or format are null.<a name="line.291"></a> +<span class="sourceLineNo">292</span> * @throws IOException<a name="line.292"></a> +<span class="sourceLineNo">293</span> * If an I/O error occurs<a name="line.293"></a> +<span class="sourceLineNo">294</span> */<a name="line.294"></a> +<span class="sourceLineNo">295</span> public static CSVParser parse(final String string, final CSVFormat format) throws IOException {<a name="line.295"></a> +<span class="sourceLineNo">296</span> Assertions.notNull(string, "string");<a name="line.296"></a> +<span class="sourceLineNo">297</span> Assertions.notNull(format, "format");<a name="line.297"></a> +<span class="sourceLineNo">298</span><a name="line.298"></a> +<span class="sourceLineNo">299</span> return new CSVParser(new StringReader(string), format);<a name="line.299"></a> +<span class="sourceLineNo">300</span> }<a name="line.300"></a> +<span class="sourceLineNo">301</span><a name="line.301"></a> +<span class="sourceLineNo">302</span> // the following objects are shared to reduce garbage<a name="line.302"></a> +<span class="sourceLineNo">303</span><a name="line.303"></a> +<span class="sourceLineNo">304</span> /**<a name="line.304"></a> +<span class="sourceLineNo">305</span> * Creates a parser for the given URL.<a name="line.305"></a> +<span class="sourceLineNo">306</span> *<a name="line.306"></a> +<span class="sourceLineNo">307</span> * <p><a name="line.307"></a> +<span class="sourceLineNo">308</span> * If you do not read all records from the given {@code url}, you should call {@link #close()} on the parser, unless<a name="line.308"></a> +<span class="sourceLineNo">309</span> * you close the {@code url}.<a name="line.309"></a> +<span class="sourceLineNo">310</span> * </p><a name="line.310"></a> +<span class="sourceLineNo">311</span> *<a name="line.311"></a> +<span class="sourceLineNo">312</span> * @param url<a name="line.312"></a> +<span class="sourceLineNo">313</span> * a URL. Must not be null.<a name="line.313"></a> +<span class="sourceLineNo">314</span> * @param charset<a name="line.314"></a> +<span class="sourceLineNo">315</span> * the charset for the resource. Must not be null.<a name="line.315"></a> +<span class="sourceLineNo">316</span> * @param format<a name="line.316"></a> +<span class="sourceLineNo">317</span> * the CSVFormat used for CSV parsing. Must not be null.<a name="line.317"></a> +<span class="sourceLineNo">318</span> * @return a new parser<a name="line.318"></a> +<span class="sourceLineNo">319</span> * @throws IllegalArgumentException<a name="line.319"></a> +<span class="sourceLineNo">320</span> * If the parameters of the format are inconsistent or if either url, charset or format are null.<a name="line.320"></a> +<span class="sourceLineNo">321</span> * @throws IOException<a name="line.321"></a> +<span class="sourceLineNo">322</span> * If an I/O error occurs<a name="line.322"></a> +<span class="sourceLineNo">323</span> */<a name="line.323"></a> +<span class="sourceLineNo">324</span> public static CSVParser parse(final URL url, final Charset charset, final CSVFormat format) throws IOException {<a name="line.324"></a> +<span class="sourceLineNo">325</span> Assertions.notNull(url, "url");<a name="line.325"></a> +<span class="sourceLineNo">326</span> Assertions.notNull(charset, "charset");<a name="line.326"></a> +<span class="sourceLineNo">327</span> Assertions.notNull(format, "format");<a name="line.327"></a> +<span class="sourceLineNo">328</span><a name="line.328"></a> +<span class="sourceLineNo">329</span> return new CSVParser(new InputStreamReader(url.openStream(), charset), format);<a name="line.329"></a> +<span class="sourceLineNo">330</span> }<a name="line.330"></a> +<span class="sourceLineNo">331</span><a name="line.331"></a> +<span class="sourceLineNo">332</span> private final CSVFormat format;<a name="line.332"></a> +<span class="sourceLineNo">333</span><a name="line.333"></a> +<span class="sourceLineNo">334</span> /** A mapping of column names to column indices */<a name="line.334"></a> +<span class="sourceLineNo">335</span> private final Map<String, Integer> headerMap;<a name="line.335"></a> +<span class="sourceLineNo">336</span><a name="line.336"></a> +<span class="sourceLineNo">337</span> /** The column order to avoid re-computing it. */<a name="line.337"></a> +<span class="sourceLineNo">338</span> private final List<String> headerNames;<a name="line.338"></a> +<span class="sourceLineNo">339</span><a name="line.339"></a> +<span class="sourceLineNo">340</span> private final Lexer lexer;<a name="line.340"></a> +<span class="sourceLineNo">341</span><a name="line.341"></a> +<span class="sourceLineNo">342</span> private final CSVRecordIterator csvRecordIterator;<a name="line.342"></a> +<span class="sourceLineNo">343</span><a name="line.343"></a> +<span class="sourceLineNo">344</span> /** A record buffer for getRecord(). Grows as necessary and is reused. */<a name="line.344"></a> +<span class="sourceLineNo">345</span> private final List<String> recordList = new ArrayList<>();<a name="line.345"></a> +<span class="sourceLineNo">346</span><a name="line.346"></a> +<span class="sourceLineNo">347</span> /**<a name="line.347"></a> +<span class="sourceLineNo">348</span> * The next record number to assign.<a name="line.348"></a> +<span class="sourceLineNo">349</span> */<a name="line.349"></a> +<span class="sourceLineNo">350</span> private long recordNumber;<a name="line.350"></a> +<span class="sourceLineNo">351</span><a name="line.351"></a> +<span class="sourceLineNo">352</span> /**<a name="line.352"></a> +<span class="sourceLineNo">353</span> * Lexer offset when the parser does not start parsing at the beginning of the source. Usually used in combination<a name="line.353"></a> +<span class="sourceLineNo">354</span> * with {@link #recordNumber}.<a name="line.354"></a> +<span class="sourceLineNo">355</span> */<a name="line.355"></a> +<span class="sourceLineNo">356</span> private final long characterOffset;<a name="line.356"></a> +<span class="sourceLineNo">357</span><a name="line.357"></a> +<span class="sourceLineNo">358</span> private final Token reusableToken = new Token();<a name="line.358"></a> +<span class="sourceLineNo">359</span><a name="line.359"></a> +<span class="sourceLineNo">360</span> /**<a name="line.360"></a> +<span class="sourceLineNo">361</span> * Customized CSV parser using the given {@link CSVFormat}<a name="line.361"></a> +<span class="sourceLineNo">362</span> *<a name="line.362"></a> +<span class="sourceLineNo">363</span> * <p><a name="line.363"></a> +<span class="sourceLineNo">364</span> * If you do not read all records from the given {@code reader}, you should call {@link #close()} on the parser,<a name="line.364"></a> +<span class="sourceLineNo">365</span> * unless you close the {@code reader}.<a name="line.365"></a> +<span class="sourceLineNo">366</span> * </p><a name="line.366"></a> +<span class="sourceLineNo">367</span> *<a name="line.367"></a> +<span class="sourceLineNo">368</span> * @param reader<a name="line.368"></a> +<span class="sourceLineNo">369</span> * a Reader containing CSV-formatted input. Must not be null.<a name="line.369"></a> +<span class="sourceLineNo">370</span> * @param format<a name="line.370"></a> +<span class="sourceLineNo">371</span> * the CSVFormat used for CSV parsing. Must not be null.<a name="line.371"></a> +<span class="sourceLineNo">372</span> * @throws IllegalArgumentException<a name="line.372"></a> +<span class="sourceLineNo">373</span> * If the parameters of the format are inconsistent or if either reader or format are null.<a name="line.373"></a> +<span class="sourceLineNo">374</span> * @throws IOException<a name="line.374"></a> +<span class="sourceLineNo">375</span> * If there is a problem reading the header or skipping the first record<a name="line.375"></a> +<span class="sourceLineNo">376</span> */<a name="line.376"></a> +<span class="sourceLineNo">377</span> public CSVParser(final Reader reader, final CSVFormat format) throws IOException {<a name="line.377"></a> +<span class="sourceLineNo">378</span> this(reader, format, 0, 1);<a name="line.378"></a> +<span class="sourceLineNo">379</span> }<a name="line.379"></a> +<span class="sourceLineNo">380</span><a name="line.380"></a> +<span class="sourceLineNo">381</span> /**<a name="line.381"></a> +<span class="sourceLineNo">382</span> * Customized CSV parser using the given {@link CSVFormat}<a name="line.382"></a> +<span class="sourceLineNo">383</span> *<a name="line.383"></a> +<span class="sourceLineNo">384</span> * <p><a name="line.384"></a> +<span class="sourceLineNo">385</span> * If you do not read all records from the given {@code reader}, you should call {@link #close()} on the parser,<a name="line.385"></a> +<span class="sourceLineNo">386</span> * unless you close the {@code reader}.<a name="line.386"></a> +<span class="sourceLineNo">387</span> * </p><a name="line.387"></a> +<span class="sourceLineNo">388</span> *<a name="line.388"></a> +<span class="sourceLineNo">389</span> * @param reader<a name="line.389"></a> +<span class="sourceLineNo">390</span> * a Reader containing CSV-formatted input. Must not be null.<a name="line.390"></a> +<span class="sourceLineNo">391</span> * @param format<a name="line.391"></a> +<span class="sourceLineNo">392</span> * the CSVFormat used for CSV parsing. Must not be null.<a name="line.392"></a> +<span class="sourceLineNo">393</span> * @param characterOffset<a name="line.393"></a> +<span class="sourceLineNo">394</span> * Lexer offset when the parser does not start parsing at the beginning of the source.<a name="line.394"></a> +<span class="sourceLineNo">395</span> * @param recordNumber<a name="line.395"></a> +<span class="sourceLineNo">396</span> * The next record number to assign<a name="line.396"></a> +<span class="sourceLineNo">397</span> * @throws IllegalArgumentException<a name="line.397"></a> +<span class="sourceLineNo">398</span> * If the parameters of the format are inconsistent or if either reader or format are null.<a name="line.398"></a> +<span class="sourceLineNo">399</span> * @throws IOException<a name="line.399"></a> +<span class="sourceLineNo">400</span> * If there is a problem reading the header or skipping the first record<a name="line.400"></a> +<span class="sourceLineNo">401</span> * @since 1.1<a name="line.401"></a> +<span class="sourceLineNo">402</span> */<a name="line.402"></a> +<span class="sourceLineNo">403</span> @SuppressWarnings("resource")<a name="line.403"></a> +<span class="sourceLineNo">404</span> public CSVParser(final Reader reader, final CSVFormat format, final long characterOffset, final long recordNumber)<a name="line.404"></a> +<span class="sourceLineNo">405</span> throws IOException {<a name="line.405"></a> +<span class="sourceLineNo">406</span> Assertions.notNull(reader, "reader");<a name="line.406"></a> +<span class="sourceLineNo">407</span> Assertions.notNull(format, "format");<a name="line.407"></a> +<span class="sourceLineNo">408</span><a name="line.408"></a> +<span class="sourceLineNo">409</span> this.format = format;<a name="line.409"></a> +<span class="sourceLineNo">410</span> this.lexer = new Lexer(format, new ExtendedBufferedReader(reader));<a name="line.410"></a> +<span class="sourceLineNo">411</span> this.csvRecordIterator = new CSVRecordIterator();<a name="line.411"></a> +<span class="sourceLineNo">412</span> final Headers headers = createHeaders();<a name="line.412"></a> +<span class="sourceLineNo">413</span> this.headerMap = headers.headerMap;<a name="line.413"></a> +<span class="sourceLineNo">414</span> this.headerNames = headers.headerNames;<a name="line.414"></a> +<span class="sourceLineNo">415</span> this.characterOffset = characterOffset;<a name="line.415"></a> +<span class="sourceLineNo">416</span> this.recordNumber = recordNumber - 1;<a name="line.416"></a> +<span class="sourceLineNo">417</span> }<a name="line.417"></a> +<span class="sourceLineNo">418</span><a name="line.418"></a> +<span class="sourceLineNo">419</span> private void addRecordValue(final boolean lastRecord) {<a name="line.419"></a> +<span class="sourceLineNo">420</span> final String input = this.reusableToken.content.toString();<a name="line.420"></a> +<span class="sourceLineNo">421</span> final String inputClean = this.format.getTrim() ? input.trim() : input;<a name="line.421"></a> +<span class="sourceLineNo">422</span> if (lastRecord && inputClean.isEmpty() && this.format.getTrailingDelimiter()) {<a name="line.422"></a> +<span class="sourceLineNo">423</span> return;<a name="line.423"></a> +<span class="sourceLineNo">424</span> }<a name="line.424"></a> +<span class="sourceLineNo">425</span> final String nullString = this.format.getNullString();<a name="line.425"></a> +<span class="sourceLineNo">426</span> this.recordList.add(inputClean.equals(nullString) ? null : inputClean);<a name="line.426"></a> +<span class="sourceLineNo">427</span> }<a name="line.427"></a> +<span class="sourceLineNo">428</span><a name="line.428"></a> +<span class="sourceLineNo">429</span> /**<a name="line.429"></a> +<span class="sourceLineNo">430</span> * Closes resources.<a name="line.430"></a> +<span class="sourceLineNo">431</span> *<a name="line.431"></a> +<span class="sourceLineNo">432</span> * @throws IOException<a name="line.432"></a> +<span class="sourceLineNo">433</span> * If an I/O error occurs<a name="line.433"></a> +<span class="sourceLineNo">434</span> */<a name="line.434"></a> +<span class="sourceLineNo">435</span> @Override<a name="line.435"></a> +<span class="sourceLineNo">436</span> public void close() throws IOException {<a name="line.436"></a> +<span class="sourceLineNo">437</span> if (this.lexer != null) {<a name="line.437"></a> +<span class="sourceLineNo">438</span> this.lexer.close();<a name="line.438"></a> +<span class="sourceLineNo">439</span> }<a name="line.439"></a> +<span class="sourceLineNo">440</span> }<a name="line.440"></a> +<span class="sourceLineNo">441</span><a name="line.441"></a> +<span class="sourceLineNo">442</span> private Map<String, Integer> createEmptyHeaderMap() {<a name="line.442"></a> +<span class="sourceLineNo">443</span> return this.format.getIgnoreHeaderCase() ?<a name="line.443"></a> +<span class="sourceLineNo">444</span> new TreeMap<>(String.CASE_INSENSITIVE_ORDER) :<a name="line.444"></a> +<span class="sourceLineNo">445</span> new LinkedHashMap<>();<a name="line.445"></a> +<span class="sourceLineNo">446</span> }<a name="line.446"></a> +<span class="sourceLineNo">447</span><a name="line.447"></a> +<span class="sourceLineNo">448</span> /**<a name="line.448"></a> +<span class="sourceLineNo">449</span> * Header information based on name and position.<a name="line.449"></a> +<span class="sourceLineNo">450</span> */<a name="line.450"></a> +<span class="sourceLineNo">451</span> private static final class Headers {<a name="line.451"></a> +<span class="sourceLineNo">452</span> /**<a name="line.452"></a> +<span class="sourceLineNo">453</span> * Header column positions (0-based)<a name="line.453"></a> +<span class="sourceLineNo">454</span> */<a name="line.454"></a> +<span class="sourceLineNo">455</span> final Map<String, Integer> headerMap;<a name="line.455"></a> +<span class="sourceLineNo">456</span><a name="line.456"></a> +<span class="sourceLineNo">457</span> /**<a name="line.457"></a> +<span class="sourceLineNo">458</span> * Header names in column order<a name="line.458"></a> +<span class="sourceLineNo">459</span> */<a name="line.459"></a> +<span class="sourceLineNo">460</span> final List<String> headerNames;<a name="line.460"></a> +<span class="sourceLineNo">461</span><a name="line.461"></a> +<span class="sourceLineNo">462</span> Headers(final Map<String, Integer> headerMap, final List<String> headerNames) {<a name="line.462"></a> +<span class="sourceLineNo">463</span> this.headerMap = headerMap;<a name="line.463"></a> +<span class="sourceLineNo">464</span> this.headerNames = headerNames;<a name="line.464"></a> +<span class="sourceLineNo">465</span> }<a name="line.465"></a> +<span class="sourceLineNo">466</span> }<a name="line.466"></a> +<span class="sourceLineNo">467</span><a name="line.467"></a> +<span class="sourceLineNo">468</span> /**<a name="line.468"></a> +<span class="sourceLineNo">469</span> * Creates the name to index mapping if the format defines a header.<a name="line.469"></a> +<span class="sourceLineNo">470</span> *<a name="line.470"></a> +<span class="sourceLineNo">471</span> * @return null if the format has no header.<a name="line.471"></a> +<span class="sourceLineNo">472</span> * @throws IOException if there is a problem reading the header or skipping the first record<a name="line.472"></a> +<span class="sourceLineNo">473</span> */<a name="line.473"></a> +<span class="sourceLineNo">474</span> private Headers createHeaders() throws IOException {<a name="line.474"></a> +<span class="sourceLineNo">475</span> Map<String, Integer> hdrMap = null;<a name="line.475"></a> +<span class="sourceLineNo">476</span> List<String> headerNames = null;<a name="line.476"></a> +<span class="sourceLineNo">477</span> final String[] formatHeader = this.format.getHeader();<a name="line.477"></a> +<span class="sourceLineNo">478</span> if (formatHeader != null) {<a name="line.478"></a> +<span class="sourceLineNo">479</span> hdrMap = createEmptyHeaderMap();<a name="line.479"></a> +<span class="sourceLineNo">480</span> String[] headerRecord = null;<a name="line.480"></a> +<span class="sourceLineNo">481</span> if (formatHeader.length == 0) {<a name="line.481"></a> +<span class="sourceLineNo">482</span> // read the header from the first line of the file<a name="line.482"></a> +<span class="sourceLineNo">483</span> final CSVRecord nextRecord = this.nextRecord();<a name="line.483"></a> +<span class="sourceLineNo">484</span> if (nextRecord != null) {<a name="line.484"></a> +<span class="sourceLineNo">485</span> headerRecord = nextRecord.values();<a name="line.485"></a> +<span class="sourceLineNo">486</span> }<a name="line.486"></a> +<span class="sourceLineNo">487</span> } else {<a name="line.487"></a> +<span class="sourceLineNo">488</span> if (this.format.getSkipHeaderRecord()) {<a name="line.488"></a> +<span class="sourceLineNo">489</span> this.nextRecord();<a name="line.489"></a> +<span class="sourceLineNo">490</span> }<a name="line.490"></a> +<span class="sourceLineNo">491</span> headerRecord = formatHeader;<a name="line.491"></a> +<span class="sourceLineNo">492</span> }<a name="line.492"></a> +<span class="sourceLineNo">493</span><a name="line.493"></a> +<span class="sourceLineNo">494</span> // build the name to index mappings<a name="line.494"></a> +<span class="sourceLineNo">495</span> if (headerRecord != null) {<a name="line.495"></a> +<span class="sourceLineNo">496</span> for (int i = 0; i < headerRecord.length; i++) {<a name="line.496"></a> +<span class="sourceLineNo">497</span> final String header = headerRecord[i];<a name="line.497"></a> +<span class="sourceLineNo">498</span> final boolean containsHeader = header == null ? false : hdrMap.containsKey(header);<a name="line.498"></a> +<span class="sourceLineNo">499</span> final boolean emptyHeader = header == null || header.trim().isEmpty();<a name="line.499"></a> +<span class="sourceLineNo">500</span> if (containsHeader) {<a name="line.500"></a> +<span class="sourceLineNo">501</span> if (!emptyHeader && !this.format.getAllowDuplicateHeaderNames()) {<a name="line.501"></a> +<span class="sourceLineNo">502</span> throw new IllegalArgumentException(<a name="line.502"></a> +<span class="sourceLineNo">503</span> String.format(<a name="line.503"></a> +<span class="sourceLineNo">504</span> "The header contains a duplicate name: \"%s\" in %s. If this is valid then use CSVFormat.withAllowDuplicateHeaderNames().",<a name="line.504"></a> +<span class="sourceLineNo">505</span> header, Arrays.toString(headerRecord)));<a name="line.505"></a> +<span class="sourceLineNo">506</span> }<a name="line.506"></a> +<span class="sourceLineNo">507</span> if (emptyHeader && !this.format.getAllowMissingColumnNames()) {<a name="line.507"></a> +<span class="sourceLineNo">508</span> throw new IllegalArgumentException(<a name="line.508"></a> +<span class="sourceLineNo">509</span> "A header name is missing in " + Arrays.toString(headerRecord));<a name="line.509"></a> +<span class="sourceLineNo">510</span> }<a name="line.510"></a> +<span class="sourceLineNo">511</span> }<a name="line.511"></a> +<span class="sourceLineNo">512</span> if (header != null) {<a name="line.512"></a> +<span class="sourceLineNo">513</span> hdrMap.put(header, Integer.valueOf(i));<a name="line.513"></a> +<span class="sourceLineNo">514</span> if (headerNames == null) {<a name="line.514"></a> +<span class="sourceLineNo">515</span> headerNames = new ArrayList<>(headerRecord.length);<a name="line.515"></a> +<span class="sourceLineNo">516</span> }<a name="line.516"></a> +<span class="sourceLineNo">517</span> headerNames.add(header);<a name="line.517"></a> +<span class="sourceLineNo">518</span> }<a name="line.518"></a> +<span class="sourceLineNo">519</span> }<a name="line.519"></a> +<span class="sourceLineNo">520</span> }<a name="line.520"></a> +<span class="sourceLineNo">521</span> }<a name="line.521"></a> +<span class="sourceLineNo">522</span> if (headerNames == null) {<a name="line.522"></a> +<span class="sourceLineNo">523</span> headerNames = Collections.emptyList(); //immutable<a name="line.523"></a> +<span class="sourceLineNo">524</span> } else {<a name="line.524"></a> +<span class="sourceLineNo">525</span> headerNames = Collections.unmodifiableList(headerNames);<a name="line.525"></a> +<span class="sourceLineNo">526</span> }<a name="line.526"></a> +<span class="sourceLineNo">527</span> return new Headers(hdrMap, headerNames);<a name="line.527"></a> +<span class="sourceLineNo">528</span> }<a name="line.528"></a> +<span class="sourceLineNo">529</span><a name="line.529"></a> +<span class="sourceLineNo">530</span> /**<a name="line.530"></a> +<span class="sourceLineNo">531</span> * Returns the current line number in the input stream.<a name="line.531"></a> +<span class="sourceLineNo">532</span> *<a name="line.532"></a> +<span class="sourceLineNo">533</span> * <p><a name="line.533"></a> +<span class="sourceLineNo">534</span> * <strong>ATTENTION:</strong> If your CSV input has multi-line values, the returned number does not correspond to<a name="line.534"></a> +<span class="sourceLineNo">535</span> * the record number.<a name="line.535"></a> +<span class="sourceLineNo">536</span> * </p><a name="line.536"></a> +<span class="sourceLineNo">537</span> *<a name="line.537"></a> +<span class="sourceLineNo">538</span> * @return current line number<a name="line.538"></a> +<span class="sourceLineNo">539</span> */<a name="line.539"></a> +<span class="sourceLineNo">540</span> public long getCurrentLineNumber() {<a name="line.540"></a> +<span class="sourceLineNo">541</span> return this.lexer.getCurrentLineNumber();<a name="line.541"></a> +<span class="sourceLineNo">542</span> }<a name="line.542"></a> +<span class="sourceLineNo">543</span><a name="line.543"></a> +<span class="sourceLineNo">544</span> /**<a name="line.544"></a> +<span class="sourceLineNo">545</span> * Gets the first end-of-line string encountered.<a name="line.545"></a> +<span class="sourceLineNo">546</span> *<a name="line.546"></a> +<span class="sourceLineNo">547</span> * @return the first end-of-line string<a name="line.547"></a> +<span class="sourceLineNo">548</span> * @since 1.5<a name="line.548"></a> +<span class="sourceLineNo">549</span> */<a name="line.549"></a> +<span class="sourceLineNo">550</span> public String getFirstEndOfLine() {<a name="line.550"></a> +<span class="sourceLineNo">551</span> return lexer.getFirstEol();<a name="line.551"></a> +<span class="sourceLineNo">552</span> }<a name="line.552"></a> +<span class="sourceLineNo">553</span><a name="line.553"></a> +<span class="sourceLineNo">554</span> /**<a name="line.554"></a> +<span class="sourceLineNo">555</span> * Returns a copy of the header map.<a name="line.555"></a> +<span class="sourceLineNo">556</span> * <p><a name="line.556"></a> +<span class="sourceLineNo">557</span> * The map keys are column names. The map values are 0-based indices.<a name="line.557"></a> +<span class="sourceLineNo">558</span> * </p><a name="line.558"></a> +<span class="sourceLineNo">559</span> * @return a copy of the header map.<a name="line.559"></a> +<span class="sourceLineNo">560</span> */<a name="line.560"></a> +<span class="sourceLineNo">561</span> public Map<String, Integer> getHeaderMap() {<a name="line.561"></a> +<span class="sourceLineNo">562</span> if (this.headerMap == null) {<a name="line.562"></a> +<span class="sourceLineNo">563</span> return null;<a name="line.563"></a> +<span class="sourceLineNo">564</span> }<a name="line.564"></a> +<span class="sourceLineNo">565</span> final Map<String, Integer> map = createEmptyHeaderMap();<a name="line.565"></a> +<span class="sourceLineNo">566</span> map.putAll(this.headerMap);<a name="line.566"></a> +<span class="sourceLineNo">567</span> return map;<a name="line.567"></a> +<span class="sourceLineNo">568</span> }<a name="line.568"></a> +<span class="sourceLineNo">569</span><a name="line.569"></a> +<span class="sourceLineNo">570</span> /**<a name="line.570"></a> +<span class="sourceLineNo">571</span> * Returns the header map.<a name="line.571"></a> +<span class="sourceLineNo">572</span> *<a name="line.572"></a> +<span class="sourceLineNo">573</span> * @return the header map.<a name="line.573"></a> +<span class="sourceLineNo">574</span> */<a name="line.574"></a> +<span class="sourceLineNo">575</span> Map<String, Integer> getHeaderMapRaw() {<a name="line.575"></a> +<span class="sourceLineNo">576</span> return this.headerMap;<a name="line.576"></a> +<span class="sourceLineNo">577</span> }<a name="line.577"></a> +<span class="sourceLineNo">578</span><a name="line.578"></a> +<span class="sourceLineNo">579</span> /**<a name="line.579"></a> +<span class="sourceLineNo">580</span> * Returns a read-only list of header names that iterates in column order.<a name="line.580"></a> +<span class="sourceLineNo">581</span> *<a name="line.581"></a> +<span class="sourceLineNo">582</span> * @return read-only list of header names that iterates in column order.<a name="line.582"></a> +<span class="sourceLineNo">583</span> * @since 1.7<a name="line.583"></a> +<span class="sourceLineNo">584</span> */<a name="line.584"></a> +<span class="sourceLineNo">585</span> public List<String> getHeaderNames() {<a name="line.585"></a> +<span class="sourceLineNo">586</span> return headerNames;<a name="line.586"></a> +<span class="sourceLineNo">587</span> }<a name="line.587"></a> +<span class="sourceLineNo">588</span><a name="line.588"></a> +<span class="sourceLineNo">589</span> /**<a name="line.589"></a> +<span class="sourceLineNo">590</span> * Returns the current record number in the input stream.<a name="line.590"></a> +<span class="sourceLineNo">591</span> *<a name="line.591"></a> +<span class="sourceLineNo">592</span> * <p><a name="line.592"></a> +<span class="sourceLineNo">593</span> * <strong>ATTENTION:</strong> If your CSV input has multi-line values, the returned number does not correspond to<a name="line.593"></a> +<span class="sourceLineNo">594</span> * the line number.<a name="line.594"></a> +<span class="sourceLineNo">595</span> * </p><a name="line.595"></a> +<span class="sourceLineNo">596</span> *<a name="line.596"></a> +<span class="sourceLineNo">597</span> * @return current record number<a name="line.597"></a> +<span class="sourceLineNo">598</span> */<a name="line.598"></a> +<span class="sourceLineNo">599</span> public long getRecordNumber() {<a name="line.599"></a> +<span class="sourceLineNo">600</span> return this.recordNumber;<a name="line.600"></a> +<span class="sourceLineNo">601</span> }<a name="line.601"></a> +<span class="sourceLineNo">602</span><a name="line.602"></a> +<span class="sourceLineNo">603</span> /**<a name="line.603"></a> +<span class="sourceLineNo">604</span> * Parses the CSV input according to the given format and returns the content as a list of<a name="line.604"></a> +<span class="sourceLineNo">605</span> * {@link CSVRecord CSVRecords}.<a name="line.605"></a> +<span class="sourceLineNo">606</span> *<a name="line.606"></a> +<span class="sourceLineNo">607</span> * <p><a name="line.607"></a> +<span class="sourceLineNo">608</span> * The returned content starts at the current parse-position in the stream.<a name="line.608"></a> +<span class="sourceLineNo">609</span> * </p><a name="line.609"></a> +<span class="sourceLineNo">610</span> *<a name="line.610"></a> +<span class="sourceLineNo">611</span> * @return list of {@link CSVRecord CSVRecords}, may be empty<a name="line.611"></a> +<span class="sourceLineNo">612</span> * @throws IOException<a name="line.612"></a> +<span class="sourceLineNo">613</span> * on parse error or input read-failure<a name="line.613"></a> +<span class="sourceLineNo">614</span> */<a name="line.614"></a> +<span class="sourceLineNo">615</span> public List<CSVRecord> getRecords() throws IOException {<a name="line.615"></a> +<span class="sourceLineNo">616</span> CSVRecord rec;<a name="line.616"></a> +<span class="sourceLineNo">617</span> final List<CSVRecord> records = new ArrayList<>();<a name="line.617"></a> +<span class="sourceLineNo">618</span> while ((rec = this.nextRecord()) != null) {<a name="line.618"></a> +<span class="sourceLineNo">619</span> records.add(rec);<a name="line.619"></a> +<span class="sourceLineNo">620</span> }<a name="line.620"></a> +<span class="sourceLineNo">621</span> return records;<a name="line.621"></a> +<span class="sourceLineNo">622</span> }<a name="line.622"></a> +<span class="sourceLineNo">623</span><a name="line.623"></a> +<span class="sourceLineNo">624</span> /**<a name="line.624"></a> +<span class="sourceLineNo">625</span> * Gets whether this parser is closed.<a name="line.625"></a> +<span class="sourceLineNo">626</span> *<a name="line.626"></a> +<span class="sourceLineNo">627</span> * @return whether this parser is closed.<a name="line.627"></a> +<span class="sourceLineNo">628</span> */<a name="line.628"></a> +<span class="sourceLineNo">629</span> public boolean isClosed() {<a name="line.629"></a> +<span class="sourceLineNo">630</span> return this.lexer.isClosed();<a name="line.630"></a> +<span class="sourceLineNo">631</span> }<a name="line.631"></a> +<span class="sourceLineNo">632</span><a name="line.632"></a> +<span class="sourceLineNo">633</span> /**<a name="line.633"></a> +<span class="sourceLineNo">634</span> * Returns an iterator on the records.<a name="line.634"></a> +<span class="sourceLineNo">635</span> *<a name="line.635"></a> +<span class="sourceLineNo">636</span> * <p><a name="line.636"></a> +<span class="sourceLineNo">637</span> * An {@link IOException} caught during the iteration are re-thrown as an<a name="line.637"></a> +<span class="sourceLineNo">638</span> * {@link IllegalStateException}.<a name="line.638"></a> +<span class="sourceLineNo">639</span> * </p><a name="line.639"></a> +<span class="sourceLineNo">640</span> * <p><a name="line.640"></a> +<span class="sourceLineNo">641</span> * If the parser is closed a call to {@link Iterator#next()} will throw a<a name="line.641"></a> +<span class="sourceLineNo">642</span> * {@link NoSuchElementException}.<a name="line.642"></a> +<span class="sourceLineNo">643</span> * </p><a name="line.643"></a> +<span class="sourceLineNo">644</span> */<a name="line.644"></a> +<span class="sourceLineNo">645</span> @Override<a name="line.645"></a> +<span class="sourceLineNo">646</span> public Iterator<CSVRecord> iterator() {<a name="line.646"></a> +<span class="sourceLineNo">647</span> return csvRecordIterator;<a name="line.647"></a> +<span class="sourceLineNo">648</span> }<a name="line.648"></a> +<span class="sourceLineNo">649</span><a name="line.649"></a> +<span class="sourceLineNo">650</span> /**<a name="line.650"></a> +<span class="sourceLineNo">651</span> * Parses the next record from the current point in the stream.<a name="line.651"></a> +<span class="sourceLineNo">652</span> *<a name="line.652"></a> +<span class="sourceLineNo">653</span> * @return the record as an array of values, or {@code null} if the end of the stream has been reached<a name="line.653"></a> +<span class="sourceLineNo">654</span> * @throws IOException<a name="line.654"></a> +<span class="sourceLineNo">655</span> * on parse error or input read-failure<a name="line.655"></a> +<span class="sourceLineNo">656</span> */<a name="line.656"></a> +<span class="sourceLineNo">657</span> CSVRecord nextRecord() throws IOException {<a name="line.657"></a> +<span class="sourceLineNo">658</span> CSVRecord result = null;<a name="line.658"></a> +<span class="sourceLineNo">659</span> this.recordList.clear();<a name="line.659"></a> +<span class="sourceLineNo">660</span> StringBuilder sb = null;<a name="line.660"></a> +<span class="sourceLineNo">661</span> final long startCharPosition = lexer.getCharacterPosition() + this.characterOffset;<a name="line.661"></a> +<span class="sourceLineNo">662</span> do {<a name="line.662"></a> +<span class="sourceLineNo">663</span> this.reusableToken.reset();<a name="line.663"></a> +<span class="sourceLineNo">664</span> this.lexer.nextToken(this.reusableToken);<a name="line.664"></a> +<span class="sourceLineNo">665</span> switch (this.reusableToken.type) {<a name="line.665"></a> +<span class="sourceLineNo">666</span> case TOKEN:<a name="line.666"></a> +<span class="sourceLineNo">667</span> this.addRecordValue(false);<a name="line.667"></a> +<span class="sourceLineNo">668</span> break;<a name="line.668"></a> +<span class="sourceLineNo">669</span> case EORECORD:<a name="line.669"></a> +<span class="sourceLineNo">670</span> this.addRecordValue(true);<a name="line.670"></a> +<span class="sourceLineNo">671</span> break;<a name="line.671"></a> +<span class="sourceLineNo">672</span> case EOF:<a name="line.672"></a> +<span class="sourceLineNo">673</span> if (this.reusableToken.isReady) {<a name="line.673"></a> +<span class="sourceLineNo">674</span> this.addRecordValue(true);<a name="line.674"></a> +<span class="sourceLineNo">675</span> }<a name="line.675"></a> +<span class="sourceLineNo">676</span> break;<a name="line.676"></a> +<span class="sourceLineNo">677</span> case INVALID:<a name="line.677"></a> +<span class="sourceLineNo">678</span> throw new IOException("(line " + this.getCurrentLineNumber() + ") invalid parse sequence");<a name="line.678"></a> +<span class="sourceLineNo">679</span> case COMMENT: // Ignored currently<a name="line.679"></a> +<span class="sourceLineNo">680</span> if (sb == null) { // first comment for this record<a name="line.680"></a> +<span class="sourceLineNo">681</span> sb = new StringBuilder();<a name="line.681"></a> +<span class="sourceLineNo">682</span> } else {<a name="line.682"></a> +<span class="sourceLineNo">683</span> sb.append(Constants.LF);<a name="line.683"></a> +<span class="sourceLineNo">684</span> }<a name="line.684"></a> +<span class="sourceLineNo">685</span> sb.append(this.reusableToken.content);<a name="line.685"></a> +<span class="sourceLineNo">686</span> this.reusableToken.type = TOKEN; // Read another token<a name="line.686"></a> +<span class="sourceLineNo">687</span> break;<a name="line.687"></a> +<span class="sourceLineNo">688</span> default:<a name="line.688"></a> +<span class="sourceLineNo">689</span> throw new IllegalStateException("Unexpected Token type: " + this.reusableToken.type);<a name="line.689"></a> +<span class="sourceLineNo">690</span> }<a name="line.690"></a> +<span class="sourceLineNo">691</span> } while (this.reusableToken.type == TOKEN);<a name="line.691"></a> +<span class="sourceLineNo">692</span><a name="line.692"></a> +<span class="sourceLineNo">693</span> if (!this.recordList.isEmpty()) {<a name="line.693"></a> +<span class="sourceLineNo">694</span> this.recordNumber++;<a name="line.694"></a> +<span class="sourceLineNo">695</span> final String comment = sb == null ? null : sb.toString();<a name="line.695"></a> +<span class="sourceLineNo">696</span> result = new CSVRecord(this, this.recordList.toArray(new String[this.recordList.size()]),<a name="line.696"></a> +<span class="sourceLineNo">697</span> comment, this.recordNumber, startCharPosition);<a name="line.697"></a> +<span class="sourceLineNo">698</span> }<a name="line.698"></a> +<span class="sourceLineNo">699</span> return result;<a name="line.699"></a> +<span class="sourceLineNo">700</span> }<a name="line.700"></a> +<span class="sourceLineNo">701</span><a name="line.701"></a> +<span class="sourceLineNo">702</span>}<a name="line.702"></a> + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +</pre> +</div> +</body> +</html>
