Menu
▾
▴
[E-p-i-c-commits] org.epic.perleditor/src/org/epic/perleditor/views/util SourceParser.java,1.15,1.16
|
From: Jochen L. <lue...@us...> - 2005-03-26 07:44:38
|
Update of /cvsroot/e-p-i-c/org.epic.perleditor/src/org/epic/perleditor/views/util In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv4005/src/org/epic/perleditor/views/util Modified Files: SourceParser.java Log Message: Performance improvements Index: SourceParser.java =================================================================== RCS file: /cvsroot/e-p-i-c/org.epic.perleditor/src/org/epic/perleditor/views/util/SourceParser.java,v retrieving revision 1.15 retrieving revision 1.16 diff -u -d -r1.15 -r1.16 --- SourceParser.java 12 Feb 2005 21:41:23 -0000 1.15 +++ SourceParser.java 26 Mar 2005 07:44:27 -0000 1.16 @@ -1,341 +1,342 @@ -package org.epic.perleditor.views.util; - -import java.util.ArrayList; -import java.util.List; -import org.eclipse.jface.text.Document; -import org.eclipse.jface.text.IDocument; -import org.eclipse.jface.text.rules.IToken; -import org.epic.perleditor.editors.perl.PerlPartitionScanner; -import org.epic.perleditor.views.model.Model; -import gnu.regexp.RE; -import gnu.regexp.REMatch; -import gnu.regexp.RESyntax; - -public class SourceParser { - - public static final String FUNCTION = "sub "; - - public static final int DO_NOT_DELETE_COMMENT_POD = 0; - public static final int DELETE_COMMENT = 1; - public static final int DELETE_POD = 2; - - /** - * line separator - */ - public static final String LINE_SEPARATOR = System - .getProperty("line.separator"); - - public SourceParser() { - //super(); - } - - public List getElements(String str, IToken tokenMatch, - String suppressString, String invalidCharacters) { - List funcList = new ArrayList(); - try { - PerlPartitionScanner scanner = new PerlPartitionScanner(); - IDocument doc = new Document(str); - scanner.setRange(doc, 0, doc.getLength()); - IToken token = scanner.nextToken(); - while (!token.isEOF()) { - if (token.equals(tokenMatch)) { - int offset = scanner.getTokenOffset(); - int length = scanner.getTokenLength(); - String foundName = str.substring(offset, offset + length); - // Check the whole linefor invalid characters - int currentLine = doc.getLineOfOffset(offset); - int lineOffset = doc.getLineOffset(currentLine); - String wholeLine = str.substring(lineOffset, lineOffset - + doc.getLineLength(currentLine)); - boolean containsInvalidCharacters = false; - char[] ic = invalidCharacters.toCharArray(); - for (int i = 0; i < ic.length; i++) { - if (wholeLine.indexOf(ic[i]) != -1) { - containsInvalidCharacters = true; - break; - } - } - if (containsInvalidCharacters) { - token = scanner.nextToken(); - continue; - } - if (foundName.endsWith(suppressString)) { - foundName = foundName.substring(0, - foundName.length() - 1); - } - // The suppressString character has to be the last character - // in the - // string. - // If not this is not a match. - // Somehow it seems that the "end sequence" of the token is - // not - // checked - // accordingly (at least not always). That's why the check - // is needed. - // This seems only to be the case for ';' as ending - // character! - else if (suppressString.equals(";")) { - token = scanner.nextToken(); - continue; - } - foundName = foundName.trim(); - length = foundName.length(); - String naked = getNaked(foundName); - if (naked != null) { - Model func = new Model(naked, offset, length); - funcList.add(func); - } - } - token = scanner.nextToken(); - } - } catch (Exception e) { - e.printStackTrace(System.err); - } - return funcList; - } - - private String getNaked(String foundName) { - if (foundName == null) { - return null; - } - if (foundName.length() <= FUNCTION.length()) { - return null; - } - foundName = foundName.trim().substring(FUNCTION.length()); - foundName = replaceInString(foundName.trim(), LINE_SEPARATOR, ""); - StringBuffer strBuf = new StringBuffer(""); - int len = foundName.length(); - boolean wasSpace = false; - for (int i = 0; i < len; i++) { - char ch = foundName.charAt(i); - if (ch == ' ') { - wasSpace = true; - } else // not space - { - if (wasSpace) { - strBuf.append(' '); - } - strBuf.append(ch); - wasSpace = false; - } - } - return strBuf.toString(); - } - - /** - * replace in a string a string sequence with another string sequence - */ - public static String replaceInString(String source, String whatBefore, - String whatAfter) { - if (null == source || source.length() == 0) { - return source; - } - int beforeLen = whatBefore.length(); - if (beforeLen == 0) { - return source; - } - StringBuffer result = new StringBuffer(""); - int lastIndex = 0; - int index = source.indexOf(whatBefore, lastIndex); - while (index >= 0) { - result.append(source.substring(lastIndex, index)); - result.append(whatAfter); - lastIndex = index + beforeLen; - // get next - index = source.indexOf(whatBefore, lastIndex); - } - result.append(source.substring(lastIndex)); - return result.toString(); - } - - /** - * Gets elements from sourcecode by using regular expressions. - * - * @param document - * @param regexp - * @param preFix - * @param postFix - * @param deleteComments - * @return - */ - public static List getElements(IDocument document, String regexp, - String preFix, String postFix, boolean deleteComments) { - String text = document.get(); - return getElements(text, regexp, preFix, postFix, deleteComments); - } - - /** - * Gets elements from sourcecode by using regular expressions. - * - * @param text - * @param regexp - * @param preFix - * @param postFix - * @param deleteComments - * @param posWordOnly - * (postioning for word-only or for complete line) - * @return - */ -public static List getElements(String text, String regexp, String preFix, - String postFix, boolean deleteComments) { - if(deleteComments == true) { - return getElements(text, regexp, preFix,postFix, DELETE_COMMENT|DELETE_POD); - } - else { - return getElements(text, regexp, preFix,postFix, DO_NOT_DELETE_COMMENT_POD); - } -} - -/** - * @param text - * @param regexp - * @param preFix - * @param postFix - * @param flags - * @return - */ -public static List getElements(String text, String regexp, String preFix, - String postFix, int flags) { - List results = new ArrayList(); - Document docOrg = new Document(text); - try { - RE reg; - // support all types of linebreaks on all platforms - //TODO it would be fine to have user-option to set for the project - // the - // line-separator - /* - * we assume a clever Linebreak-handling! (i.e. if the line-break - * could be found => the whole file consists of this kind of - * line-break - * - * otherwise, it would be a Preference-setting! (which is currently - * not available) - * - * reason for this, is that you can edit on a Windows-system - * Unix-files, which are treated within Eclipse with Unix-separtors! - * and the effect is an incorrect positioning! - */ - - RESyntax lineSep = new RESyntax(RESyntax.RE_SYNTAX_PERL5); - if (text.indexOf(System.getProperty("line.separator")) == -1) { - if (text.indexOf("\r\n") > 0) { - lineSep.setLineSeparator("\r\n"); - } else if (text.indexOf("\n") > 0) { - lineSep.setLineSeparator("\n"); - } else if (text.indexOf("\r") > 0) { - lineSep.setLineSeparator("\r"); - } - //no other possibility left for linesep, right? - } - - char[] textChar = text.toCharArray(); - - // Remove POD and comments - if ((flags&DELETE_POD) == DELETE_POD) { - REMatch[] matches; - // Remove POD - reg = new RE("^(=.*)((" + lineSep.getLineSeparator() - + ".*)+?)(" + lineSep.getLineSeparator() + "=cut)$", - RE.REG_MULTILINE, lineSep); - matches = reg.getAllMatches(text); - - for(int i=0; i < matches.length; i++) { - replaceComment(matches[i], textChar, lineSep); - } - - - } - - if ((flags&DELETE_COMMENT) == DELETE_COMMENT) { - REMatch[] matches; - // Remove comments - reg = new RE("^\\s*(#.*)$", RE.REG_MULTILINE, lineSep); - matches = reg.getAllMatches(text); - - for(int i=0; i < matches.length; i++) { - replaceComment(matches[i], textChar, lineSep); - } - - - } - - text = new String(textChar); - - reg = new RE(regexp, RE.REG_MULTILINE, lineSep); - REMatch[] matches = reg.getAllMatches(text); - for (int i = 0; i < matches.length; i++) { - String match = null; - int start; - int end; - - if (reg.getNumSubs() > 0) { - match = matches[i].toString(1); - start = matches[i].getStartIndex(1); - end = matches[i].getEndIndex(1); - } else { - match = matches[i].toString(); - start = matches[i].getStartIndex(); - end = matches[i].getEndIndex(); - } - - end -= match.length() - match.trim().length(); - match = match.trim(); - - Model func; - /* - * the following lines is for matching the word only => is - * required for Subs and Packages as well for Open Declaration - */ - func = new Model(preFix + match + postFix, start, end - start); - - results.add(func); - } - } catch (Exception e) { - e.printStackTrace(); - } - return results; - } - - /** - * @param match - * @param text - * @param reSyntax - * @return - */ - private static void replaceComment(REMatch match, char text[], RESyntax reSyntax) { -// String tmpText1 = text.substring(0, match.getStartIndex()); -// String tmpText2 = text.length() - 1 > match.getEndIndex() ? text -// .substring(match.getEndIndex() + 1) : ""; - - int start = match.getStartIndex(); - int end = match.getEndIndex(); - int lineSepLength = reSyntax.getLineSeparator().length(); - - for(int i=start; i<=end-lineSepLength; i++) { - text[i] = '#'; - } - - char[] lineSepChar = reSyntax.getLineSeparator().toCharArray(); - for(int i=0; i<lineSepChar.length; i++) { - text[end-lineSepLength+i] = lineSepChar[i]; - } - //text.replace(start, end - lineSepLength, "#"); - //text.replace(start, end - lineSepLength, "#"); - //text.replace(end - lineSepLength, end, reSyntax.getLineSeparator()); - -// int offset = text.length() - tmpText1.length() - tmpText2.length() -// - reSyntax.getLineSeparator().length(); -// if (offset > 0) { -// //we have some POD-comments here, which we now -// //replace with the same amount of chars as the text -// StringBuffer nls = new StringBuffer(); -// nls.append('#'); //First character a Comment -// nls.setLength(offset); //last character a Remark -// nls.append(reSyntax.getLineSeparator()); -// text = tmpText1 + nls + tmpText2; -// } -// -// return text; - } +package org.epic.perleditor.views.util; + +import java.util.ArrayList; +import java.util.List; +import org.eclipse.jface.text.Document; +import org.eclipse.jface.text.IDocument; +import org.eclipse.jface.text.rules.IToken; +import org.epic.perleditor.editors.perl.PerlPartitionScanner; +import org.epic.perleditor.views.model.Model; +import gnu.regexp.RE; +import gnu.regexp.REMatch; +import gnu.regexp.RESyntax; + +public class SourceParser { + + public static final String FUNCTION = "sub "; + + public static final int DO_NOT_DELETE_COMMENT_POD = 0; + public static final int DELETE_COMMENT = 1; + public static final int DELETE_POD = 2; + + /** + * line separator + */ + public static final String LINE_SEPARATOR = System + .getProperty("line.separator"); + + public SourceParser() { + //super(); + } + + public List getElements(String str, IToken tokenMatch, + String suppressString, String invalidCharacters) { + List funcList = new ArrayList(); + try { + PerlPartitionScanner scanner = new PerlPartitionScanner(); + IDocument doc = new Document(str); + scanner.setRange(doc, 0, doc.getLength()); + IToken token = scanner.nextToken(); + while (!token.isEOF()) { + if (token.equals(tokenMatch)) { + int offset = scanner.getTokenOffset(); + int length = scanner.getTokenLength(); + String foundName = str.substring(offset, offset + length); + // Check the whole linefor invalid characters + int currentLine = doc.getLineOfOffset(offset); + int lineOffset = doc.getLineOffset(currentLine); + String wholeLine = str.substring(lineOffset, lineOffset + + doc.getLineLength(currentLine)); + boolean containsInvalidCharacters = false; + char[] ic = invalidCharacters.toCharArray(); + for (int i = 0; i < ic.length; i++) { + if (wholeLine.indexOf(ic[i]) != -1) { + containsInvalidCharacters = true; + break; + } + } + if (containsInvalidCharacters) { + token = scanner.nextToken(); + continue; + } + if (foundName.endsWith(suppressString)) { + foundName = foundName.substring(0, + foundName.length() - 1); + } + // The suppressString character has to be the last character + // in the + // string. + // If not this is not a match. + // Somehow it seems that the "end sequence" of the token is + // not + // checked + // accordingly (at least not always). That's why the check + // is needed. + // This seems only to be the case for ';' as ending + // character! + else if (suppressString.equals(";")) { + token = scanner.nextToken(); + continue; + } + foundName = foundName.trim(); + length = foundName.length(); + String naked = getNaked(foundName); + if (naked != null) { + Model func = new Model(naked, offset, length); + funcList.add(func); + } + } + token = scanner.nextToken(); + } + } catch (Exception e) { + e.printStackTrace(System.err); + } + return funcList; + } + + private String getNaked(String foundName) { + if (foundName == null) { + return null; + } + if (foundName.length() <= FUNCTION.length()) { + return null; + } + foundName = foundName.trim().substring(FUNCTION.length()); + foundName = replaceInString(foundName.trim(), LINE_SEPARATOR, ""); + StringBuffer strBuf = new StringBuffer(""); + int len = foundName.length(); + boolean wasSpace = false; + for (int i = 0; i < len; i++) { + char ch = foundName.charAt(i); + if (ch == ' ') { + wasSpace = true; + } else // not space + { + if (wasSpace) { + strBuf.append(' '); + } + strBuf.append(ch); + wasSpace = false; + } + } + return strBuf.toString(); + } + + /** + * replace in a string a string sequence with another string sequence + */ + public static String replaceInString(String source, String whatBefore, + String whatAfter) { + if (null == source || source.length() == 0) { + return source; + } + int beforeLen = whatBefore.length(); + if (beforeLen == 0) { + return source; + } + StringBuffer result = new StringBuffer(""); + int lastIndex = 0; + int index = source.indexOf(whatBefore, lastIndex); + while (index >= 0) { + result.append(source.substring(lastIndex, index)); + result.append(whatAfter); + lastIndex = index + beforeLen; + // get next + index = source.indexOf(whatBefore, lastIndex); + } + result.append(source.substring(lastIndex)); + return result.toString(); + } + + /** + * Gets elements from sourcecode by using regular expressions. + * + * @param document + * @param regexp + * @param preFix + * @param postFix + * @param deleteComments + * @return + */ + public static List getElements(IDocument document, String regexp, + String preFix, String postFix, boolean deleteComments) { + String text = document.get(); + return getElements(text, regexp, preFix, postFix, deleteComments); + } + + /** + * Gets elements from sourcecode by using regular expressions. + * + * @param text + * @param regexp + * @param preFix + * @param postFix + * @param deleteComments + * @param posWordOnly + * (postioning for word-only or for complete line) + * @return + */ +public static List getElements(String text, String regexp, String preFix, + String postFix, boolean deleteComments) { + if(deleteComments == true) { + return getElements(text, regexp, preFix,postFix, DELETE_COMMENT|DELETE_POD); + } + else { + return getElements(text, regexp, preFix,postFix, DO_NOT_DELETE_COMMENT_POD); + } +} + +/** + * @param text + * @param regexp + * @param preFix + * @param postFix + * @param flags + * @return + */ +public static List getElements(String text, String regexp, String preFix, + String postFix, int flags) { + List results = new ArrayList(); + + // Document docOrg = new Document(text); + try { + RE reg; + // support all types of linebreaks on all platforms + //TODO it would be fine to have user-option to set for the project + // the + // line-separator + /* + * we assume a clever Linebreak-handling! (i.e. if the line-break + * could be found => the whole file consists of this kind of + * line-break + * + * otherwise, it would be a Preference-setting! (which is currently + * not available) + * + * reason for this, is that you can edit on a Windows-system + * Unix-files, which are treated within Eclipse with Unix-separtors! + * and the effect is an incorrect positioning! + */ + + RESyntax lineSep = new RESyntax(RESyntax.RE_SYNTAX_PERL5); + if (text.indexOf(System.getProperty("line.separator")) == -1) { + if (text.indexOf("\r\n") > 0) { + lineSep.setLineSeparator("\r\n"); + } else if (text.indexOf("\n") > 0) { + lineSep.setLineSeparator("\n"); + } else if (text.indexOf("\r") > 0) { + lineSep.setLineSeparator("\r"); + } + //no other possibility left for linesep, right? + } + + char[] textChar = text.toCharArray(); + + // Remove POD and comments + if ((flags&DELETE_POD) == DELETE_POD) { + REMatch[] matches; + // Remove POD + reg = new RE("^(=.*)((" + lineSep.getLineSeparator() + + ".*)+?)(" + lineSep.getLineSeparator() + "=cut)$", + RE.REG_MULTILINE, lineSep); + matches = reg.getAllMatches(text); + + for(int i=0; i < matches.length; i++) { + replaceComment(matches[i], textChar, lineSep); + } + + + } + + if ((flags&DELETE_COMMENT) == DELETE_COMMENT) { + REMatch[] matches; + // Remove comments + reg = new RE("^\\s*(#.*)$", RE.REG_MULTILINE, lineSep); + matches = reg.getAllMatches(text); + + for(int i=0; i < matches.length; i++) { + replaceComment(matches[i], textChar, lineSep); + } + + + } + + text = new String(textChar); + + reg = new RE(regexp, RE.REG_MULTILINE, lineSep); + REMatch[] matches = reg.getAllMatches(text); + for (int i = 0; i < matches.length; i++) { + String match = null; + int start; + int end; + + if (reg.getNumSubs() > 0) { + match = matches[i].toString(1); + start = matches[i].getStartIndex(1); + end = matches[i].getEndIndex(1); + } else { + match = matches[i].toString(); + start = matches[i].getStartIndex(); + end = matches[i].getEndIndex(); + } + + end -= match.length() - match.trim().length(); + match = match.trim(); + + Model func; + /* + * the following lines is for matching the word only => is + * required for Subs and Packages as well for Open Declaration + */ + func = new Model(preFix + match + postFix, start, end - start); + + results.add(func); + } + } catch (Exception e) { + e.printStackTrace(); + } + return results; + } + + /** + * @param match + * @param text + * @param reSyntax + * @return + */ + private static void replaceComment(REMatch match, char text[], RESyntax reSyntax) { +// String tmpText1 = text.substring(0, match.getStartIndex()); +// String tmpText2 = text.length() - 1 > match.getEndIndex() ? text +// .substring(match.getEndIndex() + 1) : ""; + + int start = match.getStartIndex(); + int end = match.getEndIndex(); + int lineSepLength = reSyntax.getLineSeparator().length(); + + for(int i=start; i<=end-lineSepLength; i++) { + text[i] = '#'; + } + + char[] lineSepChar = reSyntax.getLineSeparator().toCharArray(); + for(int i=0; i<lineSepChar.length; i++) { + text[end-lineSepLength+i] = lineSepChar[i]; + } + //text.replace(start, end - lineSepLength, "#"); + //text.replace(start, end - lineSepLength, "#"); + //text.replace(end - lineSepLength, end, reSyntax.getLineSeparator()); + +// int offset = text.length() - tmpText1.length() - tmpText2.length() +// - reSyntax.getLineSeparator().length(); +// if (offset > 0) { +// //we have some POD-comments here, which we now +// //replace with the same amount of chars as the text +// StringBuffer nls = new StringBuffer(); +// nls.append('#'); //First character a Comment +// nls.setLength(offset); //last character a Remark +// nls.append(reSyntax.getLineSeparator()); +// text = tmpText1 + nls + tmpText2; +// } +// +// return text; + } } \ No newline at end of file |