matthiasblaesing commented on issue #152: Case insensitive SHA-1 comparisons for downloaded binaries. URL: https://github.com/apache/incubator-netbeans/pull/152#issuecomment-340212245 I'm a bit ambivalent about this, as it introduces the possibility for subtle bugs. One side effect is, that if the same file is used twice in the code base it could be downloaded/cached multiple time as the hash is part of the cache file. Assuming this should be done an alternative idea would be to normalize the hash to upper case at read time. Based on current master, that would be this part in the `DownloadBinaries#execute` method: ```java String[] hashAndFile = line.split(" ", 2); if (hashAndFile.length < 2) { throw new BuildException("Bad line '" + line + "' in " + manifest, getLocation()); } if (MavenCoordinate.isMavenFile(hashAndFile[1])) { MavenCoordinate mc = MavenCoordinate.fromGradleFormat(hashAndFile[1]); fillInFile(hashAndFile[0], mc.toArtifactFilename(), manifest, () -> mavenFile(mc)); } else { fillInFile(hashAndFile[0], hashAndFile[1], manifest, () -> legacyDownload(hashAndFile[0] + "-" + hashAndFile[1])); } ``` The first entry of hashAndFile is the hash and it could be converted there. What do you think?
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services
