{"thread":{"id":"2628","subject":"Pure renames/copies","startedAt":"2005-11-21T12:01:03Z","lastAt":"2005-11-22T09:03:03Z","messageCount":11,"participants":["Santi Béjar","Linus Torvalds","Junio C Hamano","H. Peter Anvin","Santi Bejar"],"isPatch":false,"patchVersion":null,"patchTotal":null},"messages":[{"id":"12430","messageId":"87hda61a80.fsf@gmail.com","threadId":"2628","inReplyTo":null,"subject":"Pure renames/copies","fromName":"Santi Béjar","fromEmail":"sbejar@gmail.com","sentAt":"2005-11-21T12:01:03Z","receivedAt":"2005-11-21T12:01:03Z","isPatch":false,"sender":{"key":"santi@agolina.net","avatar":null},"body":"Hello:\n\n\n Is there any way to ask git to find pure renames or copies?\n\n I ask this because it is a much cheaper operation than the -C\n and -M do (-M100 does not work) and can be used when the number\n of paths if big, or when you track binary files.\n\n Thanks\n\n Santi\n"},{"id":"12456","messageId":"Pine.LNX.4.64.0511211020130.13959@g5.osdl.org","threadId":"2628","inReplyTo":"87hda61a80.fsf@gmail.com","subject":"Re: Pure renames/copies","fromName":"Linus Torvalds","fromEmail":"torvalds@osdl.org","sentAt":"2005-11-21T18:37:10Z","receivedAt":"2005-11-21T18:37:10Z","isPatch":false,"sender":{"key":"torvalds@linux-foundation.org","avatar":"https://avatars.githubusercontent.com/u/1024025?v=4"},"body":"\n\nOn Mon, 21 Nov 2005, Santi Béjar wrote:\n> \n> Is there any way to ask git to find pure renames or copies?\n\nNot directly, but git sure makes it easy for you.\n\nDo this:\n\n\tgit-diff-tree -r old..new |\n\t\tgrep '^:[^]*0000000000000000000000000000000000000000'\n\nand you'll get all the information (that \"\" is obviously the tab \ncharacter) you need to efficiently do it (or, if you want to just do one \ncommit, just do \"git-diff-tree -r cmit\"). \n\nIn the git tree, commit 0086e2c854e3af3209915e4ec2f933bcef400050 can act \nas a good example of this: the output of \n\n\tgit-diff-tree -r 0086e2c854e3af3209915e4ec2f933bcef400050\n\nis\n\n\t0086e2c854e3af3209915e4ec2f933bcef400050\n\t:100644 100644 328b399f9fe6e2b668691ab359319f50561cd773 16a8af63f0523cec82faa23f29cee579ac224e82 M .gitignore\n\t:100644 000000 a8cc5739d7851da3aeca2388d74eb92c464f1732 0000000000000000000000000000000000000000 D Documentation/git-lost+found.txt\n\t:000000 100644 0000000000000000000000000000000000000000 03156f218bb41b955779207ec2e94120f958fc45 A Documentation/git-lost-found.txt\n\t:100644 100644 a9d47c115c071694321d076af8a73a06ddd46875 1c32dd5be7156ae0e1142523fe50d84745964793 M Documentation/git.txt\n\t:100644 100644 b75cb137875b3cdb8746d2e0135e6f2743e2046a 5b2eca897386e17021d2a8a052b0c2759df96447 M Makefile\n\t:100755 000000 3892f52005d1e36676681806a87ef35dc0689f22 0000000000000000000000000000000000000000 D git-lost+found.sh\n\t:000000 100755 0000000000000000000000000000000000000000 3892f52005d1e36676681806a87ef35dc0689f22 A git-lost-found.sh\n\nand then after the \"grep\", you have just\n\n\t:100644 000000 a8cc5739d7851da3aeca2388d74eb92c464f1732 0000000000000000000000000000000000000000 D Documentation/git-lost+found.txt\n\t:000000 100644 0000000000000000000000000000000000000000 03156f218bb41b955779207ec2e94120f958fc45 A Documentation/git-lost-found.txt\n\t:100755 000000 3892f52005d1e36676681806a87ef35dc0689f22 0000000000000000000000000000000000000000 D git-lost+found.sh\n\t:000000 100755 0000000000000000000000000000000000000000 3892f52005d1e36676681806a87ef35dc0689f22 A git-lost-found.sh\n\nleft, which shows you the new and the deleted files.\n\nThen, look for renames: just match up a new file that has the same SHA1 as \na deleted file, and you can see that the change from \"git-lost+found.sh\" \nto \"git-lost-found.sh\" was exactly such an exact rename, because they \nshare the 3892f52005d1e36676681806a87ef35dc0689f22 SHA1.\n\nAfter rename detection, look at any remaining new files (in the above \nexample, only\n\n\t:000000 100644 0000000000000000000000000000000000000000 03156f218bb41b955779207ec2e94120f958fc45 A Documentation/git-lost-found.txt\n\nwould be left), and try to match up the SHA1 of that file with the result \nof \"git-ls-tree -r $old\", ie something like\n\n\tgit-ls-tree -r 0086e2c854e3af3209915e4ec2f933bcef400050^ |\n\t\tgrep 03156f218bb41b955779207ec2e94120f958fc45\n\nwhich in this case is empty (that new file wasn't an exact copy of any old \nfile, it was a rename+edit, of course).\n\nVery efficient, very simple, you can do it either with a small \nshell-script (using cut + sort + join + grep), or write a specialized tool \naround the git-diff-tree logic.\n\nOf course, arguably \"-M100\" should really do this optimization for you. \nJunio?\n\n\t\t\tLinus"},{"id":"12463","messageId":"7vu0e5re61.fsf@assigned-by-dhcp.cox.net","threadId":"2628","inReplyTo":"Pine.LNX.4.64.0511211020130.13959@g5.osdl.org","subject":"Re: Pure renames/copies","fromName":"Junio C Hamano","fromEmail":"junkio@cox.net","sentAt":"2005-11-21T19:31:18Z","receivedAt":"2005-11-21T19:31:18Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"Linus Torvalds writes:\n\n> Of course, arguably \"-M100\" should really do this optimization for you. \n> Junio?\n\nI'd agree. That is what -M100 should mean.\n"},{"id":"12464","messageId":"7vacfxrdao.fsf@assigned-by-dhcp.cox.net","threadId":"2628","inReplyTo":"Pine.LNX.4.64.0511211020130.13959@g5.osdl.org","subject":"Re: Pure renames/copies","fromName":"Junio C Hamano","fromEmail":"junkio@cox.net","sentAt":"2005-11-21T19:50:07Z","receivedAt":"2005-11-21T19:50:07Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"Linus Torvalds writes:\n\n> Of course, arguably \"-M100\" should really do this optimization for you. \n> Junio?\n\nProbably something like this would suffice.\n\n-- >8 --\nSubject: rename detection with -M100 means \"exact renames only\".\n\nWhen the user is interested in pure renames, there is no point\ndoing the similarity scores. This changes the score argument\nparsing to special case -M100 (otherwise, it is a precision\nscaled value 0 <= v < 1 and would mean 0.1, not 1.0 --- if you\ndo mean 0.1, you can say -M1), and optimizes the diffcore_rename\ntransformation to only look at pure renames in that case.\n\nSigned-off-by: Junio C Hamano \n\n---\n\ndiff --git a/diff.c b/diff.c\nindex 0391e8c..0f839c1 100644\n--- a/diff.c\n+++ b/diff.c\n@@ -853,6 +853,10 @@ static int parse_num(const char **cp_p)\n \t}\n \t*cp_p = cp;\n \n+\t/* special case: -M100 would mean 1.0 not 0.1 */\n+\tif (num == 100 && scale == 1000)\n+\t\treturn MAX_SCORE;\n+\n \t/* user says num divided by scale and we say internally that\n \t * is MAX_SCORE * num / scale.\n \t */\ndiff --git a/diffcore-rename.c b/diffcore-rename.c\nindex 6a9d95d..dba965c 100644\n--- a/diffcore-rename.c\n+++ b/diffcore-rename.c\n@@ -307,6 +307,9 @@ void diffcore_rename(struct diff_options\n \tif (rename_count == rename_dst_nr)\n \t\tgoto cleanup;\n \n+\tif (minimum_score == MAX_SCORE)\n+\t\tgoto cleanup;\n+\n \tnum_create = (rename_dst_nr - rename_count);\n \tnum_src = rename_src_nr;\n \tmx = xmalloc(sizeof(*mx) * num_create * num_src);\n"},{"id":"12471","messageId":"438235AA.8070805@zytor.com","threadId":"2628","inReplyTo":"7vacfxrdao.fsf@assigned-by-dhcp.cox.net","subject":"Re: Pure renames/copies","fromName":"H. Peter Anvin","fromEmail":"hpa@zytor.com","sentAt":"2005-11-21T21:01:30Z","receivedAt":"2005-11-21T21:01:30Z","isPatch":false,"sender":{"key":"hpa@zytor.com","avatar":null},"body":"Junio C Hamano wrote:\n> Linus Torvalds writes:\n> \n> \n>>Of course, arguably \"-M100\" should really do this optimization for you. \n>>Junio?\n> \n> \n> Probably something like this would suffice.\n> \n> -- >8 --\n> Subject: rename detection with -M100 means \"exact renames only\".\n> \n> When the user is interested in pure renames, there is no point\n> doing the similarity scores. This changes the score argument\n> parsing to special case -M100 (otherwise, it is a precision\n> scaled value 0 <= v < 1 and would mean 0.1, not 1.0 --- if you\n> do mean 0.1, you can say -M1), and optimizes the diffcore_rename\n> transformation to only look at pure renames in that case.\n> \n\nAny reason we can't make it take an actual decimal number, like -M1.0 or \n-M0.345? It seems odd and annoying to invent our own notation for \nfloating-point numbers, especially in userspace.\n\n\t-hpa\n"},{"id":"12476","messageId":"7vpsotofd0.fsf@assigned-by-dhcp.cox.net","threadId":"2628","inReplyTo":"438235AA.8070805@zytor.com","subject":"Re: Pure renames/copies","fromName":"Junio C Hamano","fromEmail":"junkio@cox.net","sentAt":"2005-11-21T21:33:47Z","receivedAt":"2005-11-21T21:33:47Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"\"H. Peter Anvin\" writes:\n\n> Any reason we can't make it take an actual decimal number, like -M1.0 or \n> -M0.345? It seems odd and annoying to invent our own notation for \n> floating-point numbers, especially in userspace.\n\nNo reason we \"can't\". About we \"don't\", inertia and nothing\nelse. It happened around this time.\n\n\thttp://marc.theaimsgroup.com/?l=git&m=111654149421574\n\nWe could in addition to take 0 <= x <= 1 decimal number and that\nshould be a simple patch to diff.c::parse_num().\n"},{"id":"12479","messageId":"43823E31.2050500@zytor.com","threadId":"2628","inReplyTo":"7vpsotofd0.fsf@assigned-by-dhcp.cox.net","subject":"Re: Pure renames/copies","fromName":"H. Peter Anvin","fromEmail":"hpa@zytor.com","sentAt":"2005-11-21T21:37:53Z","receivedAt":"2005-11-21T21:37:53Z","isPatch":false,"sender":{"key":"hpa@zytor.com","avatar":null},"body":"Junio C Hamano wrote:\n> \"H. Peter Anvin\" writes:\n> \n> \n>>Any reason we can't make it take an actual decimal number, like -M1.0 or \n>>-M0.345? It seems odd and annoying to invent our own notation for \n>>floating-point numbers, especially in userspace.\n> \n> \n> No reason we \"can't\". About we \"don't\", inertia and nothing\n> else. It happened around this time.\n> \n> \thttp://marc.theaimsgroup.com/?l=git&m=111654149421574\n> \n> We could in addition to take 0 <= x <= 1 decimal number and that\n> should be a simple patch to diff.c::parse_num().\n> \n\nOkay, in that post Linus suggests that -M without an argument should be \n== 100% (1.0), thus avoiding having to mess up the meaning of -M100 as \n0.100. It seems like a really odd thing to have -M100 mean something \nthat's completely out of line with the rest of the meaning.\n\n\t-hpa\n"},{"id":"12484","messageId":"7vd5ktoe52.fsf@assigned-by-dhcp.cox.net","threadId":"2628","inReplyTo":"43823E31.2050500@zytor.com","subject":"Re: Pure renames/copies","fromName":"Junio C Hamano","fromEmail":"junkio@cox.net","sentAt":"2005-11-21T22:00:09Z","receivedAt":"2005-11-21T22:00:09Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"\"H. Peter Anvin\" writes:\n\n> Okay, in that post Linus suggests that -M without an argument should be \n> == 100% (1.0), thus avoiding having to mess up the meaning of -M100 as \n> 0.100. It seems like a really odd thing to have -M100 mean something \n> that's completely out of line with the rest of the meaning.\n\nTrue, but it might be too late to change that; I suspect people\nexpect -M to do a bit more than pure renames by now.\n"},{"id":"12485","messageId":"438245CF.9030501@zytor.com","threadId":"2628","inReplyTo":"7vd5ktoe52.fsf@assigned-by-dhcp.cox.net","subject":"Re: Pure renames/copies","fromName":"H. Peter Anvin","fromEmail":"hpa@zytor.com","sentAt":"2005-11-21T22:10:23Z","receivedAt":"2005-11-21T22:10:23Z","isPatch":false,"sender":{"key":"hpa@zytor.com","avatar":null},"body":"Junio C Hamano wrote:\n> \"H. Peter Anvin\" writes:\n> \n> \n>>Okay, in that post Linus suggests that -M without an argument should be \n>>== 100% (1.0), thus avoiding having to mess up the meaning of -M100 as \n>>0.100. It seems like a really odd thing to have -M100 mean something \n>>that's completely out of line with the rest of the meaning.\n> \n> True, but it might be too late to change that; I suspect people\n> expect -M to do a bit more than pure renames by now.\n> \n\nOkay, how about the following? It lets both -M1.0 and -M100% work, \nwhile keeping everything else compatible, and avoiding artificial \nspecial cases.\n\n\t-hpa\n\n\ndiff --git a/diff.c b/diff.c\nindex 0391e8c..df62d2b 100644\n--- a/diff.c\n+++ b/diff.c\n@@ -843,11 +843,19 @@ static int parse_num(const char **cp_p)\n \n \tcnt = num = 0;\n \tscale = 1;\n-\twhile ('0' <= (ch = *cp) && ch <= '9') {\n-\t\tif (cnt++ < 5) {\n-\t\t\t/* We simply ignore more than 5 digits precision. */\n-\t\t\tscale *= 10;\n-\t\t\tnum = num * 10 + ch - '0';\n+\tfor(;;) {\n+\t\tch = *cp;\n+\t\tif ( ch == '.' ) {\n+\t\t\tscale = 1;\n+\t\t} else if ( ch == '%' ) {\n+\t\t\tscale = 100;\n+\t\t} else if ( ch >= '0' && ch <= '9' ) {\n+\t\t\tif ( scale < 100000 ) {\n+\t\t\t\tscale *= 10;\n+\t\t\t\tnum = (num*10) + (ch-'0');\n+\t\t\t}\n+\t\t} else {\n+\t\t\tbreak;\n \t\t}\n \t\tcp++;\n \t}\n@@ -856,7 +864,7 @@ static int parse_num(const char **cp_p)\n \t/* user says num divided by scale and we say internally that\n \t * is MAX_SCORE * num / scale.\n \t */\n-\treturn (MAX_SCORE * num / scale);\n+\treturn (num >= scale) ? MAX_SCORE : (MAX_SCORE * num / scale);\n }\n \n int diff_scoreopt_parse(const char *opt)\n"},{"id":"12487","messageId":"43824768.2070207@zytor.com","threadId":"2628","inReplyTo":"438245CF.9030501@zytor.com","subject":"Re: Pure renames/copies","fromName":"H. Peter Anvin","fromEmail":"hpa@zytor.com","sentAt":"2005-11-21T22:17:12Z","receivedAt":"2005-11-21T22:17:12Z","isPatch":false,"sender":{"key":"hpa@zytor.com","avatar":null},"body":"Better variant, which handles stuff like \"4.5%\" and rejects \n\"192.168.0.1\". Additionally, make sure numbers are unsigned (I'm making \nthem unsigned long just for the hell of it), to make sure that \nartificial wraparound scenarios don't cause harm.\n\n\t-hpa\n\n\n\ndiff --git a/diff.c b/diff.c\nindex 0391e8c..ffe8a55 100644\n--- a/diff.c\n+++ b/diff.c\n@@ -838,16 +838,29 @@ int diff_opt_parse(struct diff_options *\n \n static int parse_num(const char **cp_p)\n {\n-\tint num, scale, ch, cnt;\n+\tunsigned long num, scale;\n+\tint ch, dot;\n \tconst char *cp = *cp_p;\n \n-\tcnt = num = 0;\n+\tnum = 0;\n \tscale = 1;\n-\twhile ('0' <= (ch = *cp) && ch <= '9') {\n-\t\tif (cnt++ < 5) {\n-\t\t\t/* We simply ignore more than 5 digits precision. */\n-\t\t\tscale *= 10;\n-\t\t\tnum = num * 10 + ch - '0';\n+\tdot = 0;\n+\tfor(;;) {\n+\t\tch = *cp;\n+\t\tif ( !dot && ch == '.' ) {\n+\t\t\tscale = 1;\n+\t\t\tdot = 1;\n+\t\t} else if ( ch == '%' ) {\n+\t\t\tscale = dot ? scale*100 : 100;\n+\t\t\tcp++;\t/* % is always at the end */\n+\t\t\tbreak;\n+\t\t} else if ( ch >= '0' && ch <= '9' ) {\n+\t\t\tif ( scale < 100000 ) {\n+\t\t\t\tscale *= 10;\n+\t\t\t\tnum = (num*10) + (ch-'0');\n+\t\t\t}\n+\t\t} else {\n+\t\t\tbreak;\n \t\t}\n \t\tcp++;\n \t}\n@@ -856,7 +869,7 @@ static int parse_num(const char **cp_p)\n \t/* user says num divided by scale and we say internally that\n \t * is MAX_SCORE * num / scale.\n \t */\n-\treturn (MAX_SCORE * num / scale);\n+\treturn (num >= scale) ? MAX_SCORE : (MAX_SCORE * num / scale);\n }\n \n int diff_scoreopt_parse(const char *opt)\n"},{"id":"12528","messageId":"8aa486160511220103t7caf5ad2n@mail.gmail.com","threadId":"2628","inReplyTo":"7vacfxrdao.fsf@assigned-by-dhcp.cox.net","subject":"Re: Pure renames/copies","fromName":"Santi Bejar","fromEmail":"sbejar@gmail.com","sentAt":"2005-11-22T09:03:03Z","receivedAt":"2005-11-22T09:03:03Z","isPatch":false,"sender":{"key":"santi@agolina.net","avatar":null},"body":">\n> Probably something like this would suffice.\n>\n\nOk, thanks. Now the only issue with my broken repository (it does not\nhave all the blobs) is that it outputs:\n\nerror: unable to find abcde....\n\nfor all the src paths, but the result is ok.\n\nBut I can live with it.\n\nThanks\n"}]}