AndroidPlugin源码解析-(五)
文章目录
上篇提到，我们这篇研究的是如下过程, 分别是:
- createGenerateResValuesTask
- createMergeResourcesTask
- createMergeAssetsTask
接下来，我们一个一个来看
createGenerateResValuesTask
这个 task 实际执行下面的方法：
// 该folder名字为resValues
File folder = getResOutputDir();
List<Object> resolvedItems = getItems();
if (resolvedItems.isEmpty()) {
FileUtils.cleanOutputDir(folder);
} else {
ResValueGenerator generator = new ResValueGenerator(folder);
generator.addItems(getItems());
generator.generate();
}
这个任务的代码看起来很简单，主要就是 else 内的内容:
我们来看下generate方法:
ResValueGenerator generate
public void generate() throws IOException, ParserConfigurationException {
File pkgFolder = getFolderPath();
if (!pkgFolder.isDirectory()) {
if (!pkgFolder.mkdirs()) {
throw new RuntimeException("Failed to create " + pkgFolder.getAbsolutePath());
}
}
File resFile = new File(pkgFolder, "generated.xml");
DocumentBuilderFactory factory = DocumentBuilderFactory.newInstance();
factory.setNamespaceAware(true);
factory.setValidating(false);
factory.setIgnoringComments(true);
DocumentBuilder builder;
builder = factory.newDocumentBuilder();
Document document = builder.newDocument();
Node rootNode = document.createElement(TAG_RESOURCES);
document.appendChild(rootNode);
rootNode.appendChild(document.createTextNode("\n"));
rootNode.appendChild(document.createComment("Automatically generated file. DO NOT MODIFY"));
rootNode.appendChild(document.createTextNode("\n\n"));
for (Object item : mItems) {
if (item instanceof ClassField) {
ClassField field = (ClassField)item;
ResourceType type = ResourceType.getEnum(field.getType());
boolean hasResourceTag = (type != null && RESOURCES_WITH_TAGS.contains(type));
Node itemNode = document.createElement(hasResourceTag ? field.getType() : TAG_ITEM);
Attr nameAttr = document.createAttribute(ATTR_NAME);
nameAttr.setValue(field.getName());
itemNode.getAttributes().setNamedItem(nameAttr);
if (!hasResourceTag) {
Attr typeAttr = document.createAttribute(ATTR_TYPE);
typeAttr.setValue(field.getType());
itemNode.getAttributes().setNamedItem(typeAttr);
}
// 加入 translatable="false"
if (type == ResourceType.STRING) {
Attr translatable = document.createAttribute(ATTR_TRANSLATABLE);
translatable.setValue(VALUE_FALSE);
itemNode.getAttributes().setNamedItem(translatable);
}
if (!field.getValue().isEmpty()) {
itemNode.appendChild(document.createTextNode(field.getValue()));
}
rootNode.appendChild(itemNode);
} else if (item instanceof String) {
rootNode.appendChild(document.createTextNode("\n"));
rootNode.appendChild(document.createComment((String) item));
rootNode.appendChild(document.createTextNode("\n"));
}
}
String content;
try {
content = XmlPrettyPrinter.prettyPrint(document, true);
} catch (Throwable t) {
content = XmlUtils.toXml(document);
}
Files.write(content, resFile, Charsets.UTF_8);
}
这里就是在 gradle 的 buildType 中定义的 resValue，
最终会在build/generated/res/resValues/${buildTypes}/values目录下生成generated.xml文件，
这样，在 buildType 中定义的 resValue 就可以在代码中使用了。我在这里通常就是定义一些 string，没用过其他 resValue 类型，所以这里就不详细分析了。
通过generated.xml文件来对应这个代码，理解起来很容易。
createMergeResourcesTask
这个 task 实际执行下面的方法：
// 预处理VectorDrawable，如果没有，则忽略
ResourcePreprocessor preprocessor = getPreprocessor();
// this is full run, clean the previous output
File destinationDir = getOutputDir();
FileUtils.cleanOutputDir(destinationDir);
List<ResourceSet> resourceSets = getConfiguredResourceSets(preprocessor);
// create a new merger and populate it with the sets.
ResourceMerger merger = new ResourceMerger(minSdk);
// 这个resourceSets的来源是VariantConfiguration的getResourceSets方法
// 这个方法添加资源的顺序，决定了最后merge的时候保留的资源
for (ResourceSet resourceSet : resourceSets) {
resourceSet.loadFromFiles(getILogger());
merger.addDataSet(resourceSet);
}
// get the merged set and write it down.
// 当你使用的build tools版本是24.0.0 rc2的之后，可以在gradle.properties中设置android.enableAapt2=true属性使用aapt2
// 但是，不建议你在gradle-plugin:3.0.0-beta1之前版本使用，会有各种问题，那么我们这里关注Aapt1，而不是2版本。
// 当前版本的Aapt2的实现中存在各种Fix的注释
Aapt aapt =
AaptGradleFactory.make(
getBuilder(),
getCrunchPng(),
getProcess9Patch(),
variantScope,
getAaptTempDir());
MergedResourceWriter writer = new MergedResourceWriter(
destinationDir,
getPublicFile(),
getBlameLogFolder(),
preprocessor,
aapt::compile, // 处理.9png
getIncrementalFolder());
merger.mergeData(writer, false /*doCleanUp*/);
// No exception? Write the known state.
merger.writeBlobTo(getIncrementalFolder(), writer, false);
这里主要解释两个方法，即mergedResourceWriter的两个方法mergeDta和writeBlobTo
剩下的看代码中的注释即可。
MergedResourceWriter mergeDta
public void mergeData(@NonNull MergeConsumer<I> consumer, boolean doCleanUp)
throws MergingException {
// 对于MergeResources这个consumer来说，这就是一个赋值操作
consumer.start(mFactory);
try {
// get all the items keys.
Set<String> dataItemKeys = Sets.newHashSet();
for (S dataSet : mDataSets) {
// quick check on duplicates in the resource set.
dataSet.checkItems();
ListMultimap<String, I> map = dataSet.getDataMap();
dataItemKeys.addAll(map.keySet());
}
// loop on all the data items.
for (String dataItemKey : dataItemKeys) {
// 对于Resources来说，如果存在declare-styleable则需要merge
if (requiresMerge(dataItemKey)) {
// get all the available items, from the lower priority, to the higher
// priority
List<I> items = Lists.newArrayListWithExpectedSize(mDataSets.size());
for (S dataSet : mDataSets) {
// look for the resource key in the set
ListMultimap<String, I> itemMap = dataSet.getDataMap();
List<I> setItems = itemMap.get(dataItemKey);
items.addAll(setItems);
}
mergeItems(dataItemKey, items, consumer);
continue;
}
// for each items, look in the data sets, starting from the end of the list.
I previouslyWritten = null;
I toWrite = null;
/*
* We are looking for what to write/delete: the last non deleted item, and the
* previously written one.
*/
boolean foundIgnoredItem = false;
setLoop: for (int i = mDataSets.size() - 1 ; i >= 0 ; i--) {
S dataSet = mDataSets.get(i);
// look for the resource key in the set
ListMultimap<String, I> itemMap = dataSet.getDataMap();
List<I> items = itemMap.get(dataItemKey);
if (items.isEmpty()) {
continue;
}
// The list can contain at max 2 items. One touched and one deleted.
// More than one deleted means there was more than one which isn't possible
// More than one touched means there is more than one and this isn't possible.
for (int ii = items.size() - 1 ; ii >= 0 ; ii--) {
I item = items.get(ii);
if (consumer.ignoreItemInMerge(item)) {
foundIgnoredItem = true;
continue;
}
if (item.isWritten()) {
assert previouslyWritten == null;
previouslyWritten = item;
}
if (toWrite == null && !item.isRemoved()) {
toWrite = item;
}
if (toWrite != null && previouslyWritten != null) {
break setLoop;
}
}
}
// done searching, we should at least have something, unless we only
// found items that are not meant to be written (attr inside declare styleable)
assert foundIgnoredItem || previouslyWritten != null || toWrite != null;
if (toWrite != null && !filterAccept(toWrite)) {
toWrite = null;
}
//noinspection ConstantConditions
if (previouslyWritten == null && toWrite == null) {
continue;
}
// now need to handle, the type of each (single res file, multi res file), whether
// they are the same object or not, whether the previously written object was
// deleted.
if (toWrite == null) {
// nothing to write? delete only then.
assert previouslyWritten.isRemoved();
consumer.removeItem(previouslyWritten, null /*replacedBy*/);
} else if (previouslyWritten == null || previouslyWritten == toWrite) {
// easy one: new or updated res
consumer.addItem(toWrite);
} else {
// replacement of a resource by another.
// force write the new value
toWrite.setTouched();
consumer.addItem(toWrite);
// and remove the old one
consumer.removeItem(previouslyWritten, toWrite);
}
}
} finally {
consumer.end();
}
if (doCleanUp) {
// reset all states. We can't just reset the toWrite and previouslyWritten objects
// since overlayed items might have been touched as well.
// Should also clean (remove) objects that are removed.
postMergeCleanUp();
}
}
虽然看起来代码有点复杂，但是完成的任务挺清晰的，就是完成ResourceItem对象的 add 和 merge 操作，然后将ResourceItem对象添加到 consumer 内。
在consumer.end()的时候把这些 ResourceItem 对象写入/build/intermediates/merge${BuildType}Resources/下面的所有文件，也就是我们在工程中定义的 values 文件夹之内的那些文件。然后，将 cache 和真正的 value 一一对应，并写入 compile-file-map.properties 文件中。
接下来看这个任务中最后一个方法
MergedResourceWriter writeBlobTo
public void writeBlobTo(@NonNull File blobRootFolder, @NonNull MergeConsumer<I> consumer, boolean includeTimestamps)
throws MergingException {
// write "compact" blob
DocumentBuilder builder;
builder = mFactory.newDocumentBuilder();
Document document = builder.newDocument();
Node rootNode = document.createElement(NODE_MERGER);
// add the version code.
NodeUtils.addAttribute(document, rootNode, null, ATTR_VERSION, MERGE_BLOB_VERSION);
document.appendChild(rootNode);
for (S dataSet : mDataSets) {
Node dataSetNode = document.createElement(NODE_DATA_SET);
rootNode.appendChild(dataSetNode);
dataSet.appendToXml(dataSetNode, document, consumer, includeTimestamps);
}
// write merged items
// 创建<mergedItems><configuration><xxx></xxx></configuration></mergedItems>
// 这里是说明merge之后使用哪个资源文件的地方，如果发现merge之后图片不对，来看一下
writeAdditionalData(document, rootNode);
String content = XmlUtils.toXml(document);
try {
createDir(blobRootFolder);
} catch (IOException ioe) {
throw MergingException.wrapException(ioe).withFile(blobRootFolder).build();
}
File file = new File(blobRootFolder, FN_MERGER_XML);
try {
Files.write(content, file, Charsets.UTF_8);
} catch (IOException ioe) {
throw MergingException.wrapException(ioe).withFile(file).build();
}
}
这个方法就是将 dataSet 转化成一个 xml 文件，这个 xml 文件路径是build/intermediates/incremental/merge${buildTypes}Resources/merger.xml
最重要的两个方法是：
dataSet.appendToXml(dataSetNode, document, consumer, includeTimestamps);和writeAdditionalData(document, rootNode);
这里就不继续看内部的代码了，简单说一下这两个方法的作用，
dataSet.appendToXml(dataSetNode, document, consumer, includeTimestamps);方法的作用是:
按顺序创建<dataSet><source><file><xxx></xxx></file></source></dataSet>这样的 xml，可以在merger.xml中观察到。
writeAdditionalData(document, rootNode);方法的作用是:
创建<mergedItems><configuration><xxx></xxx></configuration></mergedItems>这样的 xml
这里是说明 merge 之后使用哪个资源文件的地方，可以在merger.xml中观察到。
这样，整 MergeResourcesTask 的过程就讲完了
createMergeAssetsTask
这个 task 实际执行下面的方法：
// this is full run, clean the previous output
File destinationDir = getOutputDir();
FileUtils.cleanOutputDir(destinationDir);
List<AssetSet> assetSets = getInputDirectorySets();
// create a new merger and populate it with the sets.
AssetMerger merger = new AssetMerger();
// AssetSet 和ResourceSet一样，都是DataSet的子类
// 所以，做的事和上面的createMergeResourcesTask方法中类似，不过一个是resourceSet对象，一个是AssetSet对象
for (AssetSet assetSet : assetSets) {
// set needs to be loaded.
assetSet.loadFromFiles(getILogger());
merger.addDataSet(assetSet);
}
// get the merged set and write it down.
MergedAssetWriter writer = new MergedAssetWriter(destinationDir);
merger.mergeData(writer, false /*doCleanUp*/);
// No exception? Write the known state.
merger.writeBlobTo(getIncrementalFolder(), writer, false);
大概浏览一下，会不会觉得这个方法和createMergeResourcesTask中的方法很相似。那么，理解起来应该就相对容易一些了。
同样的代码，不同子类的实现。
先讲一下mergeData，对于这个方法来说，MergedAssetWriter 中requiresMerge直接返回 false，也就是不能 merge。
MergedAssetWriter 中end方法什么也不做。
这里的addItem方法需要讲解一下，否则，看其他库中出现 gzip 包就不会太理解，比如 okhttp 这个库。
这里，会直接把 gzip 压缩过的文件还原为原始文件，同时会将该文件重命名为去掉.gz 的文件名 // e.g. foo.txt.gz to foo.txt
至于dataSet.appendToXml(dataSetNode, document, consumer, includeTimestamps);方法，作用和上面一样，在不同的文件夹下生成而已。
writeAdditionalData()方法就是啥也不做。
这样，整 MergeAssetsTask 的过程就讲完了。下一篇，我们继续
createBuildConfigTask, createApkProcessResTask和createProcessJavaResTasks的过程。
如果大家有不懂，欢迎通过留言和邮件进行交流。