ES-DE/es-app/src/scrapers/Scraper.cpp

314 lines
8.1 KiB
C++
Raw Normal View History

//
// Scraper.cpp
//
// Main scraper logic.
// Called from ScraperSearchComponent.
// Calls either GamesDBJSONScraper or ScreenScraper.
//
#include "scrapers/Scraper.h"
2017-11-01 22:21:10 +00:00
#include "FileData.h"
#include "GamesDBJSONScraper.h"
#include "ScreenScraper.h"
#include "Log.h"
#include "Settings.h"
2017-11-01 22:21:10 +00:00
#include "SystemData.h"
#include <FreeImage.h>
2017-11-01 22:21:10 +00:00
#include <fstream>
2017-11-03 00:33:08 +00:00
const std::map<std::string, generate_scraper_requests_func> scraper_request_funcs {
{ "TheGamesDB", &thegamesdb_generate_json_scraper_requests },
{ "ScreenScraper", &screenscraper_generate_scraper_requests }
2017-11-03 00:33:08 +00:00
};
std::unique_ptr<ScraperSearchHandle> startScraperSearch(const ScraperSearchParams& params)
{
const std::string& name = Settings::getInstance()->getString("Scraper");
std::unique_ptr<ScraperSearchHandle> handle(new ScraperSearchHandle());
// Check if the scraper in the settings still exists as a registered scraping source.
if (scraper_request_funcs.find(name) == scraper_request_funcs.end())
LOG(LogWarning) << "Configured scraper (" << name << ") unavailable, scraping aborted.";
else
scraper_request_funcs.at(name)(params, handle->mRequestQueue, handle->mResults);
return handle;
}
std::vector<std::string> getScraperList()
{
std::vector<std::string> list;
2017-11-11 14:56:22 +00:00
for(auto it = scraper_request_funcs.cbegin(); it != scraper_request_funcs.cend(); it++)
list.push_back(it->first);
return list;
}
bool isValidConfiguredScraper()
{
const std::string& name = Settings::getInstance()->getString("Scraper");
return scraper_request_funcs.find(name) != scraper_request_funcs.end();
}
// ScraperSearchHandle.
ScraperSearchHandle::ScraperSearchHandle()
{
setStatus(ASYNC_IN_PROGRESS);
}
void ScraperSearchHandle::update()
{
if(mStatus == ASYNC_DONE)
return;
if(!mRequestQueue.empty())
{
// A request can add more requests to the queue while running,
// so be careful with references into the queue.
auto& req = *(mRequestQueue.front());
AsyncHandleStatus status = req.status();
if(status == ASYNC_ERROR) {
// Propagate error.
setError(req.getStatusString());
// Empty our queue.
while(!mRequestQueue.empty())
mRequestQueue.pop();
return;
}
// Finished this one, see if we have any more.
if(status == ASYNC_DONE)
mRequestQueue.pop();
// Status == ASYNC_IN_PROGRESS.
}
// We finished without any errors!
if(mRequestQueue.empty()) {
setStatus(ASYNC_DONE);
return;
}
}
// ScraperRequest.
ScraperRequest::ScraperRequest(std::vector<ScraperSearchResult>& resultsWrite)
: mResults(resultsWrite)
{
}
// ScraperHttpRequest.
ScraperHttpRequest::ScraperHttpRequest(std::vector<ScraperSearchResult>&
resultsWrite, const std::string& url) : ScraperRequest(resultsWrite)
{
setStatus(ASYNC_IN_PROGRESS);
mReq = std::unique_ptr<HttpReq>(new HttpReq(url));
}
void ScraperHttpRequest::update()
{
HttpReq::Status status = mReq->status();
if(status == HttpReq::REQ_SUCCESS)
{
// If process() has an error, status will be changed to ASYNC_ERROR.
setStatus(ASYNC_DONE);
process(mReq, mResults);
return;
}
// Not ready yet.
if(status == HttpReq::REQ_IN_PROGRESS)
return;
// Everything else is some sort of error.
LOG(LogError) << "ScraperHttpRequest network error (status: " << status<< ") - "
<< mReq->getErrorMsg();
setError(mReq->getErrorMsg());
}
// Metadata resolving stuff.
std::unique_ptr<MDResolveHandle> resolveMetaDataAssets(const ScraperSearchResult& result,
const ScraperSearchParams& search)
{
return std::unique_ptr<MDResolveHandle>(new MDResolveHandle(result, search));
}
MDResolveHandle::MDResolveHandle(const ScraperSearchResult& result,
const ScraperSearchParams& search) : mResult(result)
{
if(!result.imageUrl.empty()) {
std::string ext;
// If we have a file extension returned by the scraper, then use it.
// Otherwise, try to guess it by the name of the URL, which point to an image.
if (!result.imageType.empty()) {
ext = result.imageType;
}
else {
size_t dot = result.imageUrl.find_last_of('.');
if (dot != std::string::npos)
ext = result.imageUrl.substr(dot, std::string::npos);
}
std::string imgPath = getSaveAsPath(search, "image", ext);
mFuncs.push_back(ResolvePair(downloadImageAsync(result.imageUrl, imgPath),
[this, imgPath] {
mResult.mdl.set("image", imgPath);
mResult.imageUrl = "";
}));
}
}
void MDResolveHandle::update()
{
if(mStatus == ASYNC_DONE || mStatus == ASYNC_ERROR)
return;
2017-11-11 14:56:22 +00:00
auto it = mFuncs.cbegin();
while(it != mFuncs.cend()) {
if(it->first->status() == ASYNC_ERROR) {
setError(it->first->getStatusString());
return;
}
else if(it->first->status() == ASYNC_DONE) {
it->second();
it = mFuncs.erase(it);
continue;
}
it++;
}
if(mFuncs.empty())
setStatus(ASYNC_DONE);
}
std::unique_ptr<ImageDownloadHandle> downloadImageAsync(const std::string& url,
const std::string& saveAs)
{
return std::unique_ptr<ImageDownloadHandle>(new ImageDownloadHandle(url, saveAs,
Settings::getInstance()->getInt("ScraperResizeWidth"),
Settings::getInstance()->getInt("ScraperResizeHeight")));
}
ImageDownloadHandle::ImageDownloadHandle(const std::string& url,
const std::string& path, int maxWidth, int maxHeight) : mSavePath(path),
mMaxWidth(maxWidth), mMaxHeight(maxHeight), mReq(new HttpReq(url))
{
}
void ImageDownloadHandle::update()
{
if(mReq->status() == HttpReq::REQ_IN_PROGRESS)
return;
if(mReq->status() != HttpReq::REQ_SUCCESS) {
std::stringstream ss;
ss << "Network error: " << mReq->getErrorMsg();
setError(ss.str());
return;
}
// Download is done, save it to disk.
std::ofstream stream(mSavePath, std::ios_base::out | std::ios_base::binary);
if(stream.bad()) {
setError("Failed to open image path to write. Permission error? Disk full?");
return;
}
const std::string& content = mReq->getContent();
stream.write(content.data(), content.length());
stream.close();
if(stream.bad()) {
setError("Failed to save image. Disk full?");
return;
}
// Resize it.
if(!resizeImage(mSavePath, mMaxWidth, mMaxHeight)) {
setError("Error saving resized image. Out of memory? Disk full?");
return;
}
setStatus(ASYNC_DONE);
}
// You can pass 0 for width or height to keep aspect ratio.
bool resizeImage(const std::string& path, int maxWidth, int maxHeight)
{
// Nothing to do.
if(maxWidth == 0 && maxHeight == 0)
return true;
FREE_IMAGE_FORMAT format = FIF_UNKNOWN;
FIBITMAP* image = NULL;
// Detect the filetype.
format = FreeImage_GetFileType(path.c_str(), 0);
if(format == FIF_UNKNOWN)
format = FreeImage_GetFIFFromFilename(path.c_str());
if(format == FIF_UNKNOWN) {
LOG(LogError) << "Error - could not detect filetype for image \"" << path << "\"!";
return false;
}
// Make sure we can read this filetype first, then load it.
if(FreeImage_FIFSupportsReading(format)) {
image = FreeImage_Load(format, path.c_str());
}
else {
LOG(LogError) << "Error - file format reading not supported for image \"" << path << "\"!";
return false;
}
float width = (float)FreeImage_GetWidth(image);
float height = (float)FreeImage_GetHeight(image);
if(maxWidth == 0)
maxWidth = (int)((maxHeight / height) * width);
else if(maxHeight == 0)
maxHeight = (int)((maxWidth / width) * height);
FIBITMAP* imageRescaled = FreeImage_Rescale(image, maxWidth, maxHeight, FILTER_BILINEAR);
FreeImage_Unload(image);
if(imageRescaled == NULL) {
LOG(LogError) << "Could not resize image! (not enough memory? invalid bitdepth?)";
return false;
}
2017-10-28 20:07:31 +00:00
bool saved = (FreeImage_Save(format, imageRescaled, path.c_str()) != 0);
FreeImage_Unload(imageRescaled);
if(!saved)
LOG(LogError) << "Failed to save resized image!";
return saved;
}
std::string getSaveAsPath(const ScraperSearchParams& params,
const std::string& suffix, const std::string& extension)
{
const std::string subdirectory = params.system->getName();
const std::string name = Utils::FileSystem::getStem(params.game->getPath()) + "-" + suffix;
std::string path = Utils::FileSystem::getHomePath() + "/.emulationstation/downloaded_images/";
if(!Utils::FileSystem::exists(path))
Utils::FileSystem::createDirectory(path);
path += subdirectory + "/";
if(!Utils::FileSystem::exists(path))
Utils::FileSystem::createDirectory(path);
path += name + extension;
return path;
}