📄 anvlrecord.html
字号:
<a name="166" href="#166">166</a> <em class="comment">// If we just had a CRLF, then its two CRLFs and its end of</em><a name="167" href="#167">167</a> <em class="comment">// record. We're done.</em><a name="168" href="#168">168</a> done = <strong>true</strong>;<a name="169" href="#169">169</a> } <strong>else</strong> {<a name="170" href="#170">170</a> isCRLF = <strong>true</strong>;<a name="171" href="#171">171</a> }<a name="172" href="#172">172</a> } <strong>else</strong> <strong>if</strong> (!recordStart && Character.isWhitespace(c)) {<a name="173" href="#173">173</a> <em class="comment">// Skip any whitespace at start of ANVLRecord.</em><a name="174" href="#174">174</a> <strong>continue</strong>;<a name="175" href="#175">175</a> } <strong>else</strong> {<a name="176" href="#176">176</a> <em class="comment">// Clear isCRLF flag if this character is NOT a '\r'.</em><a name="177" href="#177">177</a> <strong>if</strong> (isCRLF && !isCR((<strong>char</strong>)c)) {<a name="178" href="#178">178</a> isCRLF = false;<a name="179" href="#179">179</a> }<a name="180" href="#180">180</a> <em class="comment">// Not whitespace so start record if we haven't already.</em><a name="181" href="#181">181</a> <strong>if</strong> (!recordStart) {<a name="182" href="#182">182</a> recordStart = <strong>true</strong>;<a name="183" href="#183">183</a> }<a name="184" href="#184">184</a> }<a name="185" href="#185">185</a> baos.write(c);<a name="186" href="#186">186</a> }<a name="187" href="#187">187</a> <strong>return</strong> load(<strong>new</strong> String(baos.toByteArray(), UTF8));<a name="188" href="#188">188</a> }<a name="189" href="#189">189</a> <a name="190" href="#190">190</a> <em>/**<em>* </em></em><a name="191" href="#191">191</a> <em> * Parse passed String for an ANVL Record.</em><a name="192" href="#192">192</a> <em> * Looked at writing javacc grammer but preprocessing is required to</em><a name="193" href="#193">193</a> <em> * handle folding: See</em><a name="194" href="#194">194</a> <em> * https://javacc.dev.java.net/servlets/BrowseList?list=users&by=thread&from=56173.</em><a name="195" href="#195">195</a> <em> * Looked at Terence Parr's ANTLR. More capable. Can set lookahead count.</em><a name="196" href="#196">196</a> <em> * A value of 3 would help with folding. But its a pain defining UNICODE</em><a name="197" href="#197">197</a> <em> * grammers -- needed by ANVL -- and support seems incomplete</em><a name="198" href="#198">198</a> <em> * anyways: <a href="http://www.doc.ic.ac.uk/lab/secondyear/Antlr/lexer.html" target="alexandria_uri">http://www.doc.ic.ac.uk/lab/secondyear/Antlr/lexer.html</a>#unicode.</em><a name="199" href="#199">199</a> <em> * For now, go with the below hand-rolled parser.</em><a name="200" href="#200">200</a> <em> * @param s String with an ANVLRecord.</em><a name="201" href="#201">201</a> <em> * @return ANVLRecord parsed from passed String.</em><a name="202" href="#202">202</a> <em> * @throws IOException </em><a name="203" href="#203">203</a> <em> */</em><a name="204" href="#204">204</a> <strong>public</strong> <strong>static</strong> <a href="../../../../org/archive/util/anvl/ANVLRecord.html">ANVLRecord</a> load(<strong>final</strong> String s)<a name="205" href="#205">205</a> throws IOException {<a name="206" href="#206">206</a> <a href="../../../../org/archive/util/anvl/ANVLRecord.html">ANVLRecord</a> record = <strong>new</strong> <a href="../../../../org/archive/util/anvl/ANVLRecord.html">ANVLRecord</a>();<a name="207" href="#207">207</a> <strong>boolean</strong> inValue = false, inLabel = false, inComment = false, <a name="208" href="#208">208</a> inNewLine = false;<a name="209" href="#209">209</a> String label = <strong>null</strong>;<a name="210" href="#210">210</a> StringBuilder sb = <strong>new</strong> StringBuilder(s.length());<a name="211" href="#211">211</a> <strong>for</strong> (<strong>int</strong> i = 0; i < s.length(); i++) {<a name="212" href="#212">212</a> <strong>char</strong> c = s.<strong>char</strong>At(i);<a name="213" href="#213">213</a> <a name="214" href="#214">214</a> <em class="comment">// Assert I can do look-ahead.</em><a name="215" href="#215">215</a> <strong>if</strong> ((i + 1) > s.length()) {<a name="216" href="#216">216</a> <strong>throw</strong> <strong>new</strong> IOException(<span class="string">"Premature End-of-ANVLRecord:\n"</span> +<a name="217" href="#217">217</a> s.substring(i));<a name="218" href="#218">218</a> }<a name="219" href="#219">219</a> <a name="220" href="#220">220</a> <em class="comment">// If at LF of a CRLF, just go around again. Eat up the LF.</em><a name="221" href="#221">221</a> <strong>if</strong> (inNewLine && isLF(c)) {<a name="222" href="#222">222</a> <strong>continue</strong>;<a name="223" href="#223">223</a> }<a name="224" href="#224">224</a> <a name="225" href="#225">225</a> <em class="comment">// If we're at a CRLF and we were just on one, exit. Found Record.</em><a name="226" href="#226">226</a> <strong>if</strong> (inNewLine && isCR(c) && isLF(s.charAt(i + 1))) {<a name="227" href="#227">227</a> <strong>break</strong>;<a name="228" href="#228">228</a> }<a name="229" href="#229">229</a> <a name="230" href="#230">230</a> <em class="comment">// Check if we're on a fold inside a Value. Skip multiple white</em><a name="231" href="#231">231</a> <em class="comment">// space after CRLF. </em><a name="232" href="#232">232</a> <strong>if</strong> (inNewLine && inValue && Character.isWhitespace(c)) {<a name="233" href="#233">233</a> <strong>continue</strong>;<a name="234" href="#234">234</a> }<a name="235" href="#235">235</a> <a name="236" href="#236">236</a> <em class="comment">// Else set flag if we're at a CRLF.</em><a name="237" href="#237">237</a> inNewLine = isCR(c) && isLF(s.charAt(i + 1));<a name="238" href="#238">238</a> <a name="239" href="#239">239</a> <strong>if</strong> (inNewLine) {<a name="240" href="#240">240</a> <strong>if</strong> (inComment) {<a name="241" href="#241">241</a> inComment = false;<a name="242" href="#242">242</a> } <strong>else</strong> <strong>if</strong> (label != <strong>null</strong> && !inValue) {<a name="243" href="#243">243</a> <em class="comment">// Label only 'data element'.</em><a name="244" href="#244">244</a> record.addLabel(label);<a name="245" href="#245">245</a> label = <strong>null</strong>;<a name="246" href="#246">246</a> sb.setLength(0);<a name="247" href="#247">247</a> } <strong>else</strong> <strong>if</strong> (inValue) {<a name="248" href="#248">248</a> <em class="comment">// Assert I can do look-ahead past current CRLF.</em><a name="249" href="#249">249</a> <strong>if</strong> ((i + 3) > s.length()) {<a name="250" href="#250">250</a> <strong>throw</strong> <strong>new</strong> IOException(<span class="string">"Premature End-of-ANVLRecord "</span><a name="251" href="#251">251</a> + <span class="string">"(2):\n"</span> + s.substring(i));<a name="252" href="#252">252</a> }<a name="253" href="#253">253</a> <strong>if</strong> (!isCR(s.charAt(i + 2)) && !isLF(s.charAt(i + 3))<a name="254" href="#254">254</a> && Character.isWhitespace(s.charAt(i + 2))) {<a name="255" href="#255">255</a> <em class="comment">// Its a fold. Let it go around. But add in a CRLF and</em><a name="256" href="#256">256</a> <em class="comment">// space and do it here. We don't let CRLF fall through</em><a name="257" href="#257">257</a> <em class="comment">// to the sb.append on the end of this loop.</em><a name="258" href="#258">258</a> sb.append(CRLF);<a name="259" href="#259">259</a> sb.append(' ');<a name="260" href="#260">260</a> } <strong>else</strong> {<a name="261" href="#261">261</a> <em class="comment">// Next line is a new SubElement, a new Comment or</em><a name="262" href="#262">262</a> <em class="comment">// Label.</em><a name="263" href="#263">263</a> record.addLabelValue(label, sb.toString());<a name="264" href="#264">264</a> sb.setLength(0);<a name="265" href="#265">265</a> label = <strong>null</strong>;<a name="266" href="#266">266</a> inValue = false;<a name="267" href="#267">267</a> }<a name="268" href="#268">268</a> } <strong>else</strong> {<a name="269" href="#269">269</a> <em class="comment">// We're whitespace between label and value or whitespace</em><a name="270" href="#270">270</a> <em class="comment">// before we've figured whether label or comment.</em><a name="271" href="#271">271</a> }<a name="272" href="#272">272</a> <em class="comment">// Don't let the '\r' or CRLF through.</em><a name="273" href="#273">273</a> <strong>continue</strong>;<a name="274" href="#274">274</a> }<a name="275" href="#275">275</a> <a name="276" href="#276">276</a> <strong>if</strong> (inComment) {<a name="277" href="#277">277</a> <strong>continue</strong>;<a name="278" href="#278">278</a> } <strong>else</strong> <strong>if</strong> (inLabel) {<a name="279" href="#279">279</a> <strong>if</strong> (c == Label.COLON) {<a name="280" href="#280">280</a> label = sb.toString();<a name="281" href="#281">281</a> sb.setLength(0);<a name="282" href="#282">282</a> inLabel = false;<a name="283" href="#283">283</a> <strong>continue</strong>;<a name="284" href="#284">284</a> }<a name="285" href="#285">285</a> } <strong>else</strong> {<a name="286" href="#286">286</a> <strong>if</strong> (!inLabel && !inValue && !inComment) {<a name="287" href="#287">287</a> <em class="comment">// We have no state. Figure one.</em><a name="288" href="#288">288</a> <strong>if</strong> (Character.isWhitespace(c)) {<a name="289" href="#289">289</a> <em class="comment">// If no state, and whitespace, skip. Don't record.</em><a name="290" href="#290">290</a> <strong>continue</strong>;<a name="291" href="#291">291</a> } <strong>else</strong> <strong>if</strong> (label == <strong>null</strong> && c == '#') {<a name="292" href="#292">292</a> inComment = <strong>true</strong>;<a name="293" href="#293">293</a> <em class="comment">// Don't record comments.</em><a name="294" href="#294">294</a> <strong>continue</strong>;<a name="295" href="#295">295</a> } <strong>else</strong> <strong>if</strong> (label == <strong>null</strong>) {<a name="296" href="#296">296</a> inLabel = <strong>true</strong>;<a name="297" href="#297">297</a> } <strong>else</strong> {<a name="298" href="#298">298</a> inValue = <strong>true</strong>;<a name="299" href="#299">299</a> }<a name="300" href="#300">300</a> }<a name="301" href="#301">301</a> }<a name="302" href="#302">302</a> sb.append(c);<a name="303" href="#303">303</a> }<a name="304" href="#304">304</a> <strong>return</strong> record;<a name="305" href="#305">305</a> }<a name="306" href="#306">306</a> <a name="307" href="#307">307</a> <em>/**<em>*</em></em><a name="308" href="#308">308</a> <em> * @return Count of ANVLRecord bytes. Be careful, an empty ANVLRecord is</em><a name="309" href="#309">309</a> <em> * CRLFCRLF so is of size 4. Also, expensive, since it makes String of</em><a name="310" href="#310">310</a> <em> * the record so it can count bytes.</em><a name="311" href="#311">311</a> <em> */</em><a name="312" href="#312">312</a> <strong>public</strong> <strong>synchronized</strong> <strong>int</strong> getLength() {<a name="313" href="#313">313</a> <strong>int</strong> length = -1;<a name="314" href="#314">314</a> <strong>try</strong> {<a name="315" href="#315">315</a> length = getUTF8Bytes().length;<a name="316" href="#316">316</a> } <strong>catch</strong> (UnsupportedEncodingException e) {<a name="317" href="#317">317</a> <strong>throw</strong> <strong>new</strong> RuntimeException(e);<a name="318" href="#318">318</a> }<a name="319" href="#319">319</a> <strong>return</strong> length;<a name="320" href="#320">320</a> }<a name="321" href="#321">321</a> <a name="322" href="#322">322</a> <strong>public</strong> <strong>static</strong> <strong>boolean</strong> isCROrLF(<strong>final</strong> <strong>char</strong> c) {<a name="323" href="#323">323</a> <strong>return</strong> isCR(c) || isLF(c);<a name="324" href="#324">324</a> }<a name="325" href="#325">325</a> <a name="326" href="#326">326</a> <strong>public</strong> <strong>static</strong> <strong>boolean</strong> isCR(<strong>final</strong> <strong>char</strong> c) {<a name="327" href="#327">327</a> <strong>return</strong> c == ANVLRecord.CRLF.charAt(0);<a name="328" href="#328">328</a> }<a name="329" href="#329">329</a> <a name="330" href="#330">330</a> <strong>public</strong> <strong>static</strong> <strong>boolean</strong> isLF(<strong>final</strong> <strong>char</strong> c) {<a name="331" href="#331">331</a> <strong>return</strong> c == ANVLRecord.CRLF.charAt(1);<a name="332" href="#332">332</a> }<a name="333" href="#333">333</a> }</pre><hr/><div id="footer">This page was automatically generated by <a href="http://maven.apache.org/">Maven</a></div></body></html>
⌨️ 快捷键说明
复制代码
Ctrl + C
搜索代码
Ctrl + F
全屏模式
F11
切换主题
Ctrl + Shift + D
显示快捷键
?
增大字号
Ctrl + =
减小字号
Ctrl + -