#64855 Add filterbeforeconcat option to Concat
diff --git a/manual/Tasks/concat.html b/manual/Tasks/concat.html
index 513797c..71aba61 100644
--- a/manual/Tasks/concat.html
+++ b/manual/Tasks/concat.html
@@ -125,6 +125,15 @@
<td>No; default is <q>false</q></td>
</tr>
<tr>
+ <td>filterbeforeconcat</td>
+ <td>If this attribute is set to <q>true</q>, the task applies the filterchain to each
+ input after applying <code>fixlastline</code>. If this attribute is <q>false</q>, concat
+ will apply the filterchain only once to the already concatenated inputs. Filtering of
+ <code>header</code> and <code>footer</code> is not affected by this setting.
+ <em>Since Ant 1.10.10</em></td>
+ <td>No; default is <q>false</q></td>
+ </tr>
+ <tr>
<td>ignoreempty</td>
<td><em>Since Ant 1.8.0</em> Specifies whether or not the file specified
by <var>destfile</var> should be created if the source resource list is empty.
diff --git a/src/main/org/apache/tools/ant/taskdefs/Concat.java b/src/main/org/apache/tools/ant/taskdefs/Concat.java
index 333bad9..8a56046 100644
--- a/src/main/org/apache/tools/ant/taskdefs/Concat.java
+++ b/src/main/org/apache/tools/ant/taskdefs/Concat.java
@@ -212,27 +212,167 @@
}
/**
+ * This class reads a Reader and ensures it ends with the desired linebreak
+ * @since Ant 1.10.10
+ */
+ private final class LastLineFixingReader extends Reader {
+ private final Reader reader;
+ private int lastPos = 0;
+ private final char[] lastChars = new char[eolString.length()];
+ private boolean needAddSeparator = false;
+
+ private LastLineFixingReader(Reader reader) {
+ this.reader = reader;
+ }
+
+ /**
+ * Read a character from the current reader object. Advance
+ * to the next if the reader is finished.
+ * @return the character read, -1 for EOF on the last reader.
+ * @exception IOException - possibly thrown by the read for a reader
+ * object.
+ */
+ @Override
+ public int read() throws IOException {
+ if (needAddSeparator) {
+ if (lastPos >= eolString.length()) {
+ return -1;
+ } else {
+ return eolString.charAt(lastPos++);
+ }
+ }
+ int ch = reader.read();
+ if (ch == -1) {
+ if (isMissingEndOfLine()) {
+ needAddSeparator = true;
+ lastPos = 1;
+ return eolString.charAt(0);
+ }
+ } else {
+ addLastChar((char) ch);
+ return ch;
+ }
+ return -1;
+ }
+
+ /**
+ * Read into the buffer <code>cbuf</code>.
+ * @param cbuf The array to be read into.
+ * @param off The offset.
+ * @param len The length to read.
+ * @exception IOException - possibly thrown by the reads to the
+ * reader objects.
+ */
+ @Override
+ public int read(char[] cbuf, int off, int len)
+ throws IOException {
+
+ int amountRead = 0;
+ while (true) {
+ if (needAddSeparator) {
+ if (lastPos >= eolString.length()) {
+ break;
+ }
+ cbuf[off] = eolString.charAt(lastPos++);
+ len--;
+ off++;
+ amountRead++;
+ if (len == 0) {
+ return amountRead;
+ }
+ continue;
+ }
+ int nRead = reader.read(cbuf, off, len);
+ if (nRead == -1 || nRead == 0) {
+ if (isMissingEndOfLine()) {
+ needAddSeparator = true;
+ lastPos = 0;
+ } else {
+ break;
+ }
+ } else {
+ for (int i = nRead;
+ i > (nRead - lastChars.length);
+ --i) {
+ if (i <= 0) {
+ break;
+ }
+ addLastChar(cbuf[off + i - 1]);
+ }
+ len -= nRead;
+ off += nRead;
+ amountRead += nRead;
+ if (len == 0) {
+ return amountRead;
+ }
+ }
+ }
+ if (amountRead == 0) {
+ return -1;
+ }
+ return amountRead;
+ }
+
+ /**
+ * Close the current reader
+ */
+ @Override
+ public void close() throws IOException {
+ reader.close();
+ }
+
+ /**
+ * if checking for end of line at end of file
+ * add a character to the lastchars buffer
+ */
+ private void addLastChar(char ch) {
+ System.arraycopy(lastChars, 1, lastChars, 0, lastChars.length - 2 + 1);
+ lastChars[lastChars.length - 1] = ch;
+ }
+
+ /**
+ * return true if the lastchars buffer does
+ * not contain the line separator
+ */
+ private boolean isMissingEndOfLine() {
+ for (int i = 0; i < lastChars.length; ++i) {
+ if (lastChars[i] != eolString.charAt(i)) {
+ return true;
+ }
+ }
+ return false;
+ }
+ }
+
+ /**
* This class reads from each of the source files in turn.
* The concatenated result can then be filtered as
* a single stream.
*/
private final class MultiReader<S> extends Reader {
private Reader reader = null;
- private int lastPos = 0;
- private char[] lastChars = new char[eolString.length()];
- private boolean needAddSeparator = false;
private Iterator<S> readerSources;
private ReaderFactory<S> factory;
+ private final boolean filterBeforeConcat;
- private MultiReader(Iterator<S> readerSources, ReaderFactory<S> factory) {
+ private MultiReader(Iterator<S> readerSources, ReaderFactory<S> factory,
+ boolean filterBeforeConcat) {
this.readerSources = readerSources;
this.factory = factory;
+ this.filterBeforeConcat = filterBeforeConcat;
}
private Reader getReader() throws IOException {
if (reader == null && readerSources.hasNext()) {
reader = factory.getReader(readerSources.next());
- Arrays.fill(lastChars, (char) 0);
+ if(isFixLastLine())
+ {
+ reader = new LastLineFixingReader(reader);
+ }
+ if(filterBeforeConcat)
+ {
+ reader = getFilteredReader(reader);
+ }
}
return reader;
}
@@ -251,25 +391,11 @@
*/
@Override
public int read() throws IOException {
- if (needAddSeparator) {
- if (lastPos >= eolString.length()) {
- lastPos = 0;
- needAddSeparator = false;
- } else {
- return eolString.charAt(lastPos++);
- }
- }
while (getReader() != null) {
int ch = getReader().read();
if (ch == -1) {
nextReader();
- if (isFixLastLine() && isMissingEndOfLine()) {
- needAddSeparator = true;
- lastPos = 1;
- return eolString.charAt(0);
- }
} else {
- addLastChar((char) ch);
return ch;
}
}
@@ -289,39 +415,11 @@
throws IOException {
int amountRead = 0;
- while (getReader() != null || needAddSeparator) {
- if (needAddSeparator) {
- cbuf[off] = eolString.charAt(lastPos++);
- if (lastPos >= eolString.length()) {
- lastPos = 0;
- needAddSeparator = false;
- }
- len--;
- off++;
- amountRead++;
- if (len == 0) {
- return amountRead;
- }
- continue;
- }
+ while (getReader() != null) {
int nRead = getReader().read(cbuf, off, len);
if (nRead == -1 || nRead == 0) {
nextReader();
- if (isFixLastLine() && isMissingEndOfLine()) {
- needAddSeparator = true;
- lastPos = 0;
- }
} else {
- if (isFixLastLine()) {
- for (int i = nRead;
- i > (nRead - lastChars.length);
- --i) {
- if (i <= 0) {
- break;
- }
- addLastChar(cbuf[off + i - 1]);
- }
- }
len -= nRead;
off += nRead;
amountRead += nRead;
@@ -346,28 +444,6 @@
}
}
- /**
- * if checking for end of line at end of file
- * add a character to the lastchars buffer
- */
- private void addLastChar(char ch) {
- System.arraycopy(lastChars, 1, lastChars, 0, lastChars.length - 2 + 1);
- lastChars[lastChars.length - 1] = ch;
- }
-
- /**
- * return true if the lastchars buffer does
- * not contain the line separator
- */
- private boolean isMissingEndOfLine() {
- for (int i = 0; i < lastChars.length; ++i) {
- if (lastChars[i] != eolString.charAt(i)) {
- return true;
- }
- }
- return false;
- }
-
private boolean isFixLastLine() {
return fixLastLine && textBuffer == null;
}
@@ -386,8 +462,14 @@
result.setManagingComponent(this);
return result;
}
- Reader resourceReader = getFilteredReader(
- new MultiReader<>(c.iterator(), resourceReaderFactory));
+ Reader resourceReader;
+ if(filterBeforeConcat) {
+ resourceReader = new MultiReader<>(c.iterator(),
+ resourceReaderFactory, true);
+ } else {
+ resourceReader = getFilteredReader(
+ new MultiReader<>(c.iterator(), resourceReaderFactory, false));
+ }
Reader rdr;
if (header == null && footer == null) {
rdr = resourceReader;
@@ -416,7 +498,7 @@
}
}
rdr = new MultiReader<>(Arrays.asList(readers).iterator(),
- identityReaderFactory);
+ identityReaderFactory, false);
}
return outputEncoding == null ? new ReaderInputStream(rdr)
: new ReaderInputStream(rdr, outputEncoding);
@@ -454,6 +536,9 @@
/** Stores the binary attribute */
private boolean binary;
+ /** Stores the filterBeforeConcat attribute */
+ private boolean filterBeforeConcat;
+
// Child elements.
/**
@@ -780,6 +865,17 @@
}
/**
+ * Set the filterBeforeConcat attribute. If true, concat will filter each
+ * input through the filterchain before concatenating the results. This
+ * allows to e.g. use the FileTokenizer to tokenize each input.
+ * @param filterBeforeConcat if true, filter each input before concatenation
+ * @since Ant 1.10.10
+ */
+ public void setFilterBeforeConcat(final boolean filterBeforeConcat) {
+ this.filterBeforeConcat = filterBeforeConcat;
+ }
+
+ /**
* Execute the concat task.
*/
@Override
diff --git a/src/tests/antunit/taskdefs/concat-test.xml b/src/tests/antunit/taskdefs/concat-test.xml
index c596d90..82182ff 100644
--- a/src/tests/antunit/taskdefs/concat-test.xml
+++ b/src/tests/antunit/taskdefs/concat-test.xml
@@ -118,6 +118,50 @@
</au:assertTrue>
</target>
+ <target name="testFilterIsAppliedToConcatenation"
+ depends="-fixlastline-setup"
+ description="https://bz.apache.org/bugzilla/show_bug.cgi?id=64855">
+ <au:assertTrue>
+ <resourcesmatch>
+ <string>1${line.separator}2${line.separator}1${line.separator}2${line.separator}</string>
+ <concat fixlastline="true">
+ <filelist dir="${input}">
+ <file name="1"/>
+ <file name="2"/>
+ </filelist>
+ <filterchain>
+ <tokenfilter>
+ <filetokenizer/>
+ <replaceregex pattern="(.*)" flags="s" replace="\1\1"/>
+ </tokenfilter>
+ </filterchain>
+ </concat>
+ </resourcesmatch>
+ </au:assertTrue>
+ </target>
+
+ <target name="testFilterBeforeConcatActuallyAppliesFilterToEachInput"
+ depends="-fixlastline-setup"
+ description="https://bz.apache.org/bugzilla/show_bug.cgi?id=64855">
+ <au:assertTrue>
+ <resourcesmatch>
+ <string>1${line.separator}1${line.separator}2${line.separator}2${line.separator}</string>
+ <concat fixlastline="true" filterbeforeconcat="true">
+ <filelist dir="${input}">
+ <file name="1"/>
+ <file name="2"/>
+ </filelist>
+ <filterchain>
+ <tokenfilter>
+ <filetokenizer/>
+ <replaceregex pattern="(.*)" flags="s" replace="\1\1"/>
+ </tokenfilter>
+ </filterchain>
+ </concat>
+ </resourcesmatch>
+ </au:assertTrue>
+ </target>
+
<target name="testIgnoreEmptyFalseFileIsCreated">
<mkdir dir="${input}" />
<mkdir dir="${output}" />
