mercurial/metadata.py

copies-rust: tokenize all paths into integer

Copy information for each changesets tend to affect a small new number of path.
However, each of these path might be handled a large number of time. Handling
HgPathBuf (aka `Vec<u8>`) is expensive. Handling integer is cheap.

With this patch we:
- turn any input path into an integer "token" early,
- do all the internal logic using such "token",
- turn "token" back into path right before returning a result.

This gives use a quite significant performance boost in our slower cases.

Repo            Case                                  Source-Rev   Dest-Rev      # of revisions  old time      new time      Difference    Factor    time per rev
---------------------------------------------------------------------------------------------------------------------------------------------------------------
pypy            x000_revs_x000_added_x_copies         8a3b5bfd266e 2c68e87c3efe :   6780 revs,   0.092828 s,   0.081225 s,  -0.011603 s, × 0.8750,    11 µs/rev
pypy            x0000_revs_x_added_0_copies           d1defd0dc478 c9cb1334cc78 :  43645 revs,   0.711975 s,   0.586011 s,  -0.125964 s, × 0.8231,    13 µs/rev
pypy            x0000_revs_xx000_added_x000_copies    08ea3258278e d9fa043f30c0 :  11316 revs,   0.124505 s,   0.114173 s,  -0.010332 s, × 0.9170,    10 µs/rev
netbeans        x000_revs_x000_added_x000_copies      ff453e9fee32 411350406ec2 :   5750 revs,   0.072072 s,   0.061004 s,  -0.011068 s, × 0.8464,    10 µs/rev
netbeans        x0000_revs_xx000_added_x000_copies    588c2d1ced70 1aad62e59ddd :  66949 revs,   0.682732 s,   0.535874 s,  -0.146858 s, × 0.7849,     8 µs/rev
mozilla-central x00000_revs_x0000_added_x0000_copies  6832ae71433c 4c222a1d9a00 : 153721 revs,   1.935918 s,   1.781383 s,  -0.154535 s, × 0.9202,    11 µs/rev
mozilla-central x00000_revs_x00000_added_x000_copies  76caed42cf7c 1daa622bbe42 : 204976 revs,   2.827320 s,   2.603867 s,  -0.223453 s, × 0.9210,    12 µs/rev
mozilla-try     x0000_revs_xx000_added_x000_copies    89294cd501d9 7ccb2fc7ccb5 :  97052 revs,   3.243010 s,   1.529120 s,  -1.713890 s, × 0.4715,    15 µs/rev
mozilla-try     x00000_revs_x_added_0_copies          6a320851d377 1ebb79acd503 : 363753 revs,   5.693818 s,   4.842699 s,  -0.851119 s, × 0.8505,    13 µs/rev
mozilla-try     x00000_revs_x_added_x_copies          5173c4b6f97c 95d83ee7242d : 362229 revs,   5.677655 s,   4.761732 s,  -0.915923 s, × 0.8387,    13 µs/rev
mozilla-try     x00000_revs_x000_added_x_copies       9126823d0e9c ca82787bb23c : 359344 revs,   5.563370 s,   4.733912 s,  -0.829458 s, × 0.8509,    13 µs/rev
mozilla-try     x00000_revs_x0000_added_x0000_copies  8d3fafa80d4b eb884023b810 : 192665 revs,   2.864099 s,   2.593410 s,  -0.270689 s, × 0.9055,    13 µs/rev
mozilla-try     x00000_revs_x00000_added_x0000_copies 1b661134e2ca 1ae03d022d6d : 228985 revs, 113.297287 s,  41.041198 s, -72.256089 s, × 0.3622,   179 µs/rev
mozilla-try     x00000_revs_x00000_added_x000_copies  9b2a99adc05e 8e29777b48e6 : 382065 revs,  59.498652 s,  27.915689 s, -31.582963 s, × 0.4692,    73 µs/rev





Full timing comparison between this revision and the previous one:


Repo            Case                                  Source-Rev   Dest-Rev      # of revisions  old time      new time      Difference    Factor    time per rev
---------------------------------------------------------------------------------------------------------------------------------------------------------------
mercurial       x_revs_x_added_0_copies               ad6b123de1c7 39cfcef4f463 :      1 revs,   0.000042 s,   0.000042 s,  +0.000000 s, × 1.0000,    42 µs/rev
mercurial       x_revs_x_added_x_copies               2b1c78674230 0c1d10351869 :      6 revs,   0.000104 s,   0.000110 s,  +0.000006 s, × 1.0577,    18 µs/rev
mercurial       x000_revs_x000_added_x_copies         81f8ff2a9bf2 dd3267698d84 :   1032 revs,   0.004913 s,   0.004918 s,  +0.000005 s, × 1.0010,     4 µs/rev
pypy            x_revs_x_added_0_copies               aed021ee8ae8 099ed31b181b :      9 revs,   0.000191 s,   0.000195 s,  +0.000004 s, × 1.0209,    21 µs/rev
pypy            x_revs_x000_added_0_copies            4aa4e1f8e19a 359343b9ac0e :      1 revs,   0.000050 s,   0.000049 s,  -0.000001 s, × 0.9800,    49 µs/rev
pypy            x_revs_x_added_x_copies               ac52eb7bbbb0 72e022663155 :      7 revs,   0.000112 s,   0.000112 s,  +0.000000 s, × 1.0000,    16 µs/rev
pypy            x_revs_x00_added_x_copies             c3b14617fbd7 ace7255d9a26 :      1 revs,   0.000288 s,   0.000324 s,  +0.000036 s, × 1.1250,   324 µs/rev
pypy            x_revs_x000_added_x000_copies         df6f7a526b60 a83dc6a2d56f :      6 revs,   0.010411 s,   0.010611 s,  +0.000200 s, × 1.0192,  1768 µs/rev
pypy            x000_revs_xx00_added_0_copies         89a76aede314 2f22446ff07e :   4785 revs,   0.052852 s,   0.050835 s,  -0.002017 s, × 0.9618,    10 µs/rev
pypy            x000_revs_x000_added_x_copies         8a3b5bfd266e 2c68e87c3efe :   6780 revs,   0.092828 s,   0.081225 s,  -0.011603 s, × 0.8750,    11 µs/rev
pypy            x000_revs_x000_added_x000_copies      89a76aede314 7b3dda341c84 :   5441 revs,   0.063269 s,   0.061291 s,  -0.001978 s, × 0.9687,    11 µs/rev
pypy            x0000_revs_x_added_0_copies           d1defd0dc478 c9cb1334cc78 :  43645 revs,   0.711975 s,   0.586011 s,  -0.125964 s, × 0.8231,    13 µs/rev
pypy            x0000_revs_xx000_added_0_copies       bf2c629d0071 4ffed77c095c :      2 revs,   0.012771 s,   0.012824 s,  +0.000053 s, × 1.0042,  6412 µs/rev
pypy            x0000_revs_xx000_added_x000_copies    08ea3258278e d9fa043f30c0 :  11316 revs,   0.124505 s,   0.114173 s,  -0.010332 s, × 0.9170,    10 µs/rev
netbeans        x_revs_x_added_0_copies               fb0955ffcbcd a01e9239f9e7 :      2 revs,   0.000082 s,   0.000085 s,  +0.000003 s, × 1.0366,    42 µs/rev
netbeans        x_revs_x000_added_0_copies            6f360122949f 20eb231cc7d0 :      2 revs,   0.000111 s,   0.000108 s,  -0.000003 s, × 0.9730,    54 µs/rev
netbeans        x_revs_x_added_x_copies               1ada3faf6fb6 5a39d12eecf4 :      3 revs,   0.000171 s,   0.000175 s,  +0.000004 s, × 1.0234,    58 µs/rev
netbeans        x_revs_x00_added_x_copies             35be93ba1e2c 9eec5e90c05f :      9 revs,   0.000708 s,   0.000719 s,  +0.000011 s, × 1.0155,    79 µs/rev
netbeans        x000_revs_xx00_added_0_copies         eac3045b4fdd 51d4ae7f1290 :   1421 revs,   0.010608 s,   0.010175 s,  -0.000433 s, × 0.9592,     7 µs/rev
netbeans        x000_revs_x000_added_x_copies         e2063d266acd 6081d72689dc :   1533 revs,   0.015635 s,   0.015569 s,  -0.000066 s, × 0.9958,    10 µs/rev
netbeans        x000_revs_x000_added_x000_copies      ff453e9fee32 411350406ec2 :   5750 revs,   0.072072 s,   0.061004 s,  -0.011068 s, × 0.8464,    10 µs/rev
netbeans        x0000_revs_xx000_added_x000_copies    588c2d1ced70 1aad62e59ddd :  66949 revs,   0.682732 s,   0.535874 s,  -0.146858 s, × 0.7849,     8 µs/rev
mozilla-central x_revs_x_added_0_copies               3697f962bb7b 7015fcdd43a2 :      2 revs,   0.000090 s,   0.000090 s,  +0.000000 s, × 1.0000,    45 µs/rev
mozilla-central x_revs_x000_added_0_copies            dd390860c6c9 40d0c5bed75d :      8 revs,   0.000210 s,   0.000281 s,  +0.000071 s, × 1.3381,    35 µs/rev
mozilla-central x_revs_x_added_x_copies               8d198483ae3b 14207ffc2b2f :      9 revs,   0.000182 s,   0.000187 s,  +0.000005 s, × 1.0275,    20 µs/rev
mozilla-central x_revs_x00_added_x_copies             98cbc58cc6bc 446a150332c3 :      7 revs,   0.000594 s,   0.000660 s,  +0.000066 s, × 1.1111,    94 µs/rev
mozilla-central x_revs_x000_added_x000_copies         3c684b4b8f68 0a5e72d1b479 :      3 revs,   0.003102 s,   0.003385 s,  +0.000283 s, × 1.0912,  1128 µs/rev
mozilla-central x_revs_x0000_added_x0000_copies       effb563bb7e5 c07a39dc4e80 :      6 revs,   0.060234 s,   0.069812 s,  +0.009578 s, × 1.1590, 11635 µs/rev
mozilla-central x000_revs_xx00_added_0_copies         6100d773079a 04a55431795e :   1593 revs,   0.006300 s,   0.006503 s,  +0.000203 s, × 1.0322,     4 µs/rev
mozilla-central x000_revs_x000_added_x_copies         9f17a6fc04f9 2d37b966abed :     41 revs,   0.004817 s,   0.004988 s,  +0.000171 s, × 1.0355,   121 µs/rev
mozilla-central x000_revs_x000_added_x000_copies      7c97034feb78 4407bd0c6330 :   7839 revs,   0.065451 s,   0.063963 s,  -0.001488 s, × 0.9773,     8 µs/rev
mozilla-central x0000_revs_xx000_added_0_copies       9eec5917337d 67118cc6dcad :    615 revs,   0.026282 s,   0.026225 s,  -0.000057 s, × 0.9978,    42 µs/rev
mozilla-central x0000_revs_xx000_added_x000_copies    f78c615a656c 96a38b690156 :  30263 revs,   0.206873 s,   0.201377 s,  -0.005496 s, × 0.9734,     6 µs/rev
mozilla-central x00000_revs_x0000_added_x0000_copies  6832ae71433c 4c222a1d9a00 : 153721 revs,   1.935918 s,   1.781383 s,  -0.154535 s, × 0.9202,    11 µs/rev
mozilla-central x00000_revs_x00000_added_x000_copies  76caed42cf7c 1daa622bbe42 : 204976 revs,   2.827320 s,   2.603867 s,  -0.223453 s, × 0.9210,    12 µs/rev
mozilla-try     x_revs_x_added_0_copies               aaf6dde0deb8 9790f499805a :      2 revs,   0.000842 s,   0.000845 s,  +0.000003 s, × 1.0036,   422 µs/rev
mozilla-try     x_revs_x000_added_0_copies            d8d0222927b4 5bb8ce8c7450 :      2 revs,   0.000870 s,   0.000862 s,  -0.000008 s, × 0.9908,   431 µs/rev
mozilla-try     x_revs_x_added_x_copies               092fcca11bdb 936255a0384a :      4 revs,   0.000165 s,   0.000161 s,  -0.000004 s, × 0.9758,    40 µs/rev
mozilla-try     x_revs_x00_added_x_copies             b53d2fadbdb5 017afae788ec :      2 revs,   0.001145 s,   0.001163 s,  +0.000018 s, × 1.0157,   581 µs/rev
mozilla-try     x_revs_x000_added_x000_copies         20408ad61ce5 6f0ee96e21ad :      1 revs,   0.026500 s,   0.032414 s,  +0.005914 s, × 1.2232, 32414 µs/rev
mozilla-try     x_revs_x0000_added_x0000_copies       effb563bb7e5 c07a39dc4e80 :      6 revs,   0.059407 s,   0.070149 s,  +0.010742 s, × 1.1808, 11691 µs/rev
mozilla-try     x000_revs_xx00_added_0_copies         6100d773079a 04a55431795e :   1593 revs,   0.006325 s,   0.006526 s,  +0.000201 s, × 1.0318,     4 µs/rev
mozilla-try     x000_revs_x000_added_x_copies         9f17a6fc04f9 2d37b966abed :     41 revs,   0.005171 s,   0.005187 s,  +0.000016 s, × 1.0031,   126 µs/rev
mozilla-try     x000_revs_x000_added_x000_copies      1346fd0130e4 4c65cbdabc1f :   6657 revs,   0.066837 s,   0.065047 s,  -0.001790 s, × 0.9732,     9 µs/rev
mozilla-try     x0000_revs_x_added_0_copies           63519bfd42ee a36a2a865d92 :  40314 revs,   0.314252 s,   0.301129 s,  -0.013123 s, × 0.9582,     7 µs/rev
mozilla-try     x0000_revs_x_added_x_copies           9fe69ff0762d bcabf2a78927 :  38690 revs,   0.304160 s,   0.280683 s,  -0.023477 s, × 0.9228,     7 µs/rev
mozilla-try     x0000_revs_xx000_added_x_copies       156f6e2674f2 4d0f2c178e66 :   8598 revs,   0.089223 s,   0.084897 s,  -0.004326 s, × 0.9515,     9 µs/rev
mozilla-try     x0000_revs_xx000_added_0_copies       9eec5917337d 67118cc6dcad :    615 revs,   0.026711 s,   0.026620 s,  -0.000091 s, × 0.9966,    43 µs/rev
mozilla-try     x0000_revs_xx000_added_x000_copies    89294cd501d9 7ccb2fc7ccb5 :  97052 revs,   3.243010 s,   1.529120 s,  -1.713890 s, × 0.4715,    15 µs/rev
mozilla-try     x0000_revs_x0000_added_x0000_copies   e928c65095ed e951f4ad123a :  52031 revs,   0.756500 s,   0.738709 s,  -0.017791 s, × 0.9765,    14 µs/rev
mozilla-try     x00000_revs_x_added_0_copies          6a320851d377 1ebb79acd503 : 363753 revs,   5.693818 s,   4.842699 s,  -0.851119 s, × 0.8505,    13 µs/rev
mozilla-try     x00000_revs_x00000_added_0_copies     dc8a3ca7010e d16fde900c9c :  34414 revs,   0.590904 s,   0.596946 s,  +0.006042 s, × 1.0102,    17 µs/rev
mozilla-try     x00000_revs_x_added_x_copies          5173c4b6f97c 95d83ee7242d : 362229 revs,   5.677655 s,   4.761732 s,  -0.915923 s, × 0.8387,    13 µs/rev
mozilla-try     x00000_revs_x000_added_x_copies       9126823d0e9c ca82787bb23c : 359344 revs,   5.563370 s,   4.733912 s,  -0.829458 s, × 0.8509,    13 µs/rev
mozilla-try     x00000_revs_x0000_added_x0000_copies  8d3fafa80d4b eb884023b810 : 192665 revs,   2.864099 s,   2.593410 s,  -0.270689 s, × 0.9055,    13 µs/rev
mozilla-try     x00000_revs_x00000_added_x0000_copies 1b661134e2ca 1ae03d022d6d : 228985 revs, 113.297287 s,  41.041198 s, -72.256089 s, × 0.3622,   179 µs/rev
mozilla-try     x00000_revs_x00000_added_x000_copies  9b2a99adc05e 8e29777b48e6 : 382065 revs,  59.498652 s,  27.915689 s, -31.582963 s, × 0.4692,    73 µs/rev


Full timing comparison between this revision and the filelog copy tracing.

Repo            Case                                  Source-Rev   Dest-Rev      # of revisions   filelog      sidedata      Difference    Factor    time per rev
---------------------------------------------------------------------------------------------------------------------------------------------------------------
mercurial       x_revs_x_added_0_copies               ad6b123de1c7 39cfcef4f463 :       1 revs,   0.000903 s,  0.000042 s,  -0.000861 s, × 0.0465,    41 µs/rev
mercurial       x_revs_x_added_x_copies               2b1c78674230 0c1d10351869 :       6 revs,   0.001861 s,  0.000110 s,  -0.001751 s, × 0.0591,    18 µs/rev
mercurial       x000_revs_x000_added_x_copies         81f8ff2a9bf2 dd3267698d84 :    1032 revs,   0.018577 s,  0.004918 s,  -0.013659 s, × 0.2647,     4 µs/rev
pypy            x_revs_x_added_0_copies               aed021ee8ae8 099ed31b181b :       9 revs,   0.001519 s,  0.000195 s,  -0.001324 s, × 0.1283,    21 µs/rev
pypy            x_revs_x000_added_0_copies            4aa4e1f8e19a 359343b9ac0e :       1 revs,   0.213855 s,  0.000049 s,  -0.350d73 s, × 0.0002,    48 µs/rev
pypy            x_revs_x_added_x_copies               ac52eb7bbbb0 72e022663155 :       7 revs,   0.017022 s,  0.000112 s,  -0.016910 s, × 0.0065,    15 µs/rev
pypy            x_revs_x00_added_x_copies             c3b14617fbd7 ace7255d9a26 :       1 revs,   0.019398 s,  0.000324 s,  -0.019074 s, × 0.0167,   323 µs/rev
pypy            x_revs_x000_added_x000_copies         df6f7a526b60 a83dc6a2d56f :       6 revs,   0.769467 s,  0.010611 s,  -0.758856 s, × 0.0137,  1768 µs/rev
pypy            x000_revs_xx00_added_0_copies         89a76aede314 2f22446ff07e :    4785 revs,   1.221952 s,  0.050835 s,  -1.171117 s, × 0.0416,    10 µs/rev
pypy            x000_revs_x000_added_x_copies         8a3b5bfd266e 2c68e87c3efe :    6780 revs,   1.304007 s,  0.081225 s,  -1.222782 s, × 0.0622,    11 µs/rev
pypy            x000_revs_x000_added_x000_copies      89a76aede314 7b3dda341c84 :    5441 revs,   1.686610 s,  0.061291 s,  -1.625319 s, × 0.0363,    11 µs/rev
pypy            x0000_revs_x_added_0_copies           d1defd0dc478 c9cb1334cc78 :   43645 revs,   0.001107 s,  0.586011 s,  +0.584904 s, × 529.36,    13 µs/rev
pypy            x0000_revs_xx000_added_0_copies       bf2c629d0071 4ffed77c095c :       2 revs,   1.100760 s,  0.012824 s,  -1.087936 s, × 0.0116,  6408 µs/rev
pypy            x0000_revs_xx000_added_x000_copies    08ea3258278e d9fa043f30c0 :   11316 revs,   1.350547 s,  0.114173 s,  -1.236374 s, × 0.0845,    10 µs/rev
netbeans        x_revs_x_added_0_copies               fb0955ffcbcd a01e9239f9e7 :       2 revs,   0.027864 s,  0.000085 s,  -0.027779 s, × 0.0030,    42 µs/rev
netbeans        x_revs_x000_added_0_copies            6f360122949f 20eb231cc7d0 :       2 revs,   0.132479 s,  0.000108 s,  -0.132371 s, × 0.0008,    53 µs/rev
netbeans        x_revs_x_added_x_copies               1ada3faf6fb6 5a39d12eecf4 :       3 revs,   0.025405 s,  0.000175 s,  -0.025230 s, × 0.0068,    58 µs/rev
netbeans        x_revs_x00_added_x_copies             35be93ba1e2c 9eec5e90c05f :       9 revs,   0.053244 s,  0.000719 s,  -0.052525 s, × 0.0135,    79 µs/rev
netbeans        x000_revs_xx00_added_0_copies         eac3045b4fdd 51d4ae7f1290 :    1421 revs,   0.038017 s,  0.010175 s,  -0.027842 s, × 0.2676,     7 µs/rev
netbeans        x000_revs_x000_added_x_copies         e2063d266acd 6081d72689dc :    1533 revs,   0.198308 s,  0.015569 s,  -0.182739 s, × 0.0785,    10 µs/rev
netbeans        x000_revs_x000_added_x000_copies      ff453e9fee32 411350406ec2 :    5750 revs,   0.949749 s,  0.061004 s,  -0.888745 s, × 0.0642,    10 µs/rev
netbeans        x0000_revs_xx000_added_x000_copies    588c2d1ced70 1aad62e59ddd :   66949 revs,   3.932262 s,  0.535874 s,  -3.396388 s, × 0.1362,     8 µs/rev
mozilla-central x_revs_x_added_0_copies               3697f962bb7b 7015fcdd43a2 :       2 revs,   0.024490 s,  0.000090 s,  -0.024400 s, × 0.0036,    44 µs/rev
mozilla-central x_revs_x000_added_0_copies            dd390860c6c9 40d0c5bed75d :       8 revs,   0.143885 s,  0.000281 s,  -0.143604 s, × 0.0019,    35 µs/rev
mozilla-central x_revs_x_added_x_copies               8d198483ae3b 14207ffc2b2f :       9 revs,   0.025471 s,  0.000187 s,  -0.025284 s, × 0.0073,    20 µs/rev
mozilla-central x_revs_x00_added_x_copies             98cbc58cc6bc 446a150332c3 :       7 revs,   0.086013 s,  0.000660 s,  -0.085353 s, × 0.0076,    94 µs/rev
mozilla-central x_revs_x000_added_x000_copies         3c684b4b8f68 0a5e72d1b479 :       3 revs,   0.200726 s,  0.003385 s,  -0.197341 s, × 0.0168,  1127 µs/rev
mozilla-central x_revs_x0000_added_x0000_copies       effb563bb7e5 c07a39dc4e80 :       6 revs,   2.224171 s,  0.069812 s,  -2.154359 s, × 0.0313, 11633 µs/rev
mozilla-central x000_revs_xx00_added_0_copies         6100d773079a 04a55431795e :    1593 revs,   0.090780 s,  0.006503 s,  -0.084277 s, × 0.0716,     4 µs/rev
mozilla-central x000_revs_x000_added_x_copies         9f17a6fc04f9 2d37b966abed :      41 revs,   0.764805 s,  0.004988 s,  -0.759817 s, × 0.0065,   121 µs/rev
mozilla-central x000_revs_x000_added_x000_copies      7c97034feb78 4407bd0c6330 :    7839 revs,   1.161405 s,  0.063963 s,  -1.097442 s, × 0.0550,     8 µs/rev
mozilla-central x0000_revs_xx000_added_0_copies       9eec5917337d 67118cc6dcad :     615 revs,   6.816186 s,  0.026225 s,  -6.789961 s, × 0.0038,    42 µs/rev
mozilla-central x0000_revs_xx000_added_x000_copies    f78c615a656c 96a38b690156 :   30263 revs,   3.374819 s,  0.201377 s,  -3.173442 s, × 0.0596,     6 µs/rev
mozilla-central x00000_revs_x0000_added_x0000_copies  6832ae71433c 4c222a1d9a00 :  153721 revs,  16.285469 s,  1.781383 s, -14.504086 s, × 0.1093,    11 µs/rev
mozilla-central x00000_revs_x00000_added_x000_copies  76caed42cf7c 1daa622bbe42 :  204976 revs,  21.207733 s,  2.603867 s, -18.603866 s, × 0.1227,    12 µs/rev
mozilla-try     x_revs_x_added_0_copies               aaf6dde0deb8 9790f499805a :       2 revs,   0.080843 s,  0.000845 s,  -0.079998 s, × 0.0104,   422 µs/rev
mozilla-try     x_revs_x000_added_0_copies            d8d0222927b4 5bb8ce8c7450 :       2 revs,   0.511068 s,  0.000862 s,  -0.510206 s, × 0.0016,   430 µs/rev
mozilla-try     x_revs_x_added_x_copies               092fcca11bdb 936255a0384a :       4 revs,   0.021573 s,  0.000161 s,  -0.021412 s, × 0.0074,    40 µs/rev
mozilla-try     x_revs_x00_added_x_copies             b53d2fadbdb5 017afae788ec :       2 revs,   0.227726 s,  0.001163 s,  -0.226563 s, × 0.0051,   581 µs/rev
mozilla-try     x_revs_x000_added_x000_copies         20408ad61ce5 6f0ee96e21ad :       1 revs,   1.120448 s,  0.032414 s,  -1.088034 s, × 0.0289, 32381 µs/rev
mozilla-try     x_revs_x0000_added_x0000_copies       effb563bb7e5 c07a39dc4e80 :       6 revs,   2.241713 s,  0.070149 s,  -2.171564 s, × 0.0312, 11689 µs/rev
mozilla-try     x000_revs_xx00_added_0_copies         6100d773079a 04a55431795e :    1593 revs,   0.090633 s,  0.006526 s,  -0.084107 s, × 0.0720,     4 µs/rev
mozilla-try     x000_revs_x000_added_x_copies         9f17a6fc04f9 2d37b966abed :      41 revs,   0.770403 s,  0.005187 s,  -0.765216 s, × 0.0067,   126 µs/rev
mozilla-try     x000_revs_x000_added_x000_copies      1346fd0130e4 4c65cbdabc1f :    6657 revs,   1.184557 s,  0.065047 s,  -1.119510 s, × 0.0549,     9 µs/rev
mozilla-try     x0000_revs_x_added_0_copies           63519bfd42ee a36a2a865d92 :   40314 revs,   0.085790 s,  0.301129 s,  +0.215339 s, × 3.5100,     7 µs/rev
mozilla-try     x0000_revs_x_added_x_copies           9fe69ff0762d bcabf2a78927 :   38690 revs,   0.080616 s,  0.280683 s,  +0.200067 s, × 3.4817,     7 µs/rev
mozilla-try     x0000_revs_xx000_added_x_copies       156f6e2674f2 4d0f2c178e66 :    8598 revs,   7.712554 s,  0.084897 s,  -7.627657 s, × 0.0110,     9 µs/rev
mozilla-try     x0000_revs_xx000_added_0_copies       9eec5917337d 67118cc6dcad :     615 revs,   6.937294 s,  0.026620 s,  -6.910674 s, × 0.0038,    43 µs/rev
mozilla-try     x0000_revs_xx000_added_x000_copies    89294cd501d9 7ccb2fc7ccb5 :   97052 revs,   7.712313 s,  1.529120 s,  -6.183193 s, × 0.1982,    15 µs/rev
mozilla-try     x0000_revs_x0000_added_x0000_copies   e928c65095ed e951f4ad123a :   52031 revs,   9.966910 s,  0.738709 s,  -9.228201 s, × 0.0741,    14 µs/rev
mozilla-try     x00000_revs_x_added_0_copies          6a320851d377 1ebb79acd503 :  363753 revs,   0.090397 s,  4.842699 s,  +4.752302 s, × 53.571,    13 µs/rev
mozilla-try     x00000_revs_x00000_added_0_copies     dc8a3ca7010e d16fde900c9c :   34414 revs,  27.817167 s,  0.596946 s, -27.220221 s, × 0.0214,    17 µs/rev
mozilla-try     x00000_revs_x_added_x_copies          5173c4b6f97c 95d83ee7242d :  362229 revs,   0.091305 s,  4.761732 s,  +4.670427 s, × 52.151,    13 µs/rev
mozilla-try     x00000_revs_x000_added_x_copies       9126823d0e9c ca82787bb23c :  359344 revs,   0.231183 s,  4.733912 s,  +4.502729 s, × 20.476,    13 µs/rev
mozilla-try     x00000_revs_x0000_added_x0000_copies  8d3fafa80d4b eb884023b810 :  192665 revs,  19.830617 s,  2.593410 s, -17.237207 s, × 0.1307,    13 µs/rev
mozilla-try     x00000_revs_x00000_added_x0000_copies 1b661134e2ca 1ae03d022d6d :  228985 revs,  21.743873 s, 41.041198 s, +19.297325 s, × 1.8874,   179 µs/rev
mozilla-try     x00000_revs_x00000_added_x000_copies  9b2a99adc05e 8e29777b48e6 :  382065 revs,  25.935037 s, 27.915689 s,  +1.980652 s, × 1.0763,    73 µs/rev

Differential Revision: https://phab.mercurial-scm.org/D9493

# coding: utf-8
# metadata.py -- code related to various metadata computation and access.
#
# Copyright 2019 Google, Inc <[e-mail]>
# Copyright 2020 Pierre-Yves David <[e-mail]>
#
# This software may be used and distributed according to the terms of the
# GNU General Public License version 2 or any later version.
from __future__ import absolute_import, print_function

import multiprocessing
import struct

from .node import (
    nullid,
    nullrev,
)
from . import (
    error,
    pycompat,
    util,
)

from .revlogutils import (
    flagutil as sidedataflag,
    sidedata as sidedatamod,
)


class ChangingFiles(object):
    """A class recording the changes made to files by a changeset

    Actions performed on files are gathered into 3 sets:

    - added:   files actively added in the changeset.
    - merged:  files whose history got merged
    - removed: files removed in the revision
    - salvaged: files that might have been deleted by a merge but were not
    - touched: files affected by the merge

    and copies information is held by 2 mappings

    - copied_from_p1: {"<new-name>": "<source-name-in-p1>"} mapping for copies
    - copied_from_p2: {"<new-name>": "<source-name-in-p2>"} mapping for copies

    See their inline help for details.
    """

    def __init__(
        self,
        touched=None,
        added=None,
        removed=None,
        merged=None,
        salvaged=None,
        p1_copies=None,
        p2_copies=None,
    ):
        self._added = set(() if added is None else added)
        self._merged = set(() if merged is None else merged)
        self._removed = set(() if removed is None else removed)
        self._touched = set(() if touched is None else touched)
        self._salvaged = set(() if salvaged is None else salvaged)
        self._touched.update(self._added)
        self._touched.update(self._merged)
        self._touched.update(self._removed)
        self._p1_copies = dict(() if p1_copies is None else p1_copies)
        self._p2_copies = dict(() if p2_copies is None else p2_copies)

    def __eq__(self, other):
        return (
            self.added == other.added
            and self.merged == other.merged
            and self.removed == other.removed
            and self.salvaged == other.salvaged
            and self.touched == other.touched
            and self.copied_from_p1 == other.copied_from_p1
            and self.copied_from_p2 == other.copied_from_p2
        )

    @property
    def has_copies_info(self):
        return bool(
            self.removed
            or self.merged
            or self.salvaged
            or self.copied_from_p1
            or self.copied_from_p2
        )

    @util.propertycache
    def added(self):
        """files actively added in the changeset

        Any file present in that revision that was absent in all the changeset's
        parents.

        In case of merge, this means a file absent in one of the parents but
        existing in the other will *not* be contained in this set. (They were
        added by an ancestor)
        """
        return frozenset(self._added)

    def mark_added(self, filename):
        if 'added' in vars(self):
            del self.added
        self._added.add(filename)
        self.mark_touched(filename)

    def update_added(self, filenames):
        for f in filenames:
            self.mark_added(f)

    @util.propertycache
    def merged(self):
        """files actively merged during a merge

        Any modified files which had modification on both size that needed merging.

        In this case a new filenode was created and it has two parents.
        """
        return frozenset(self._merged)

    def mark_merged(self, filename):
        if 'merged' in vars(self):
            del self.merged
        self._merged.add(filename)
        self.mark_touched(filename)

    def update_merged(self, filenames):
        for f in filenames:
            self.mark_merged(f)

    @util.propertycache
    def removed(self):
        """files actively removed by the changeset

        In case of merge this will only contain the set of files removing "new"
        content. For any file absent in the current changeset:

        a) If the file exists in both parents, it is clearly "actively" removed
        by this changeset.

        b) If a file exists in only one parent and in none of the common
        ancestors, then the file was newly added in one of the merged branches
        and then got "actively" removed.

        c) If a file exists in only one parent and at least one of the common
        ancestors using the same filenode, then the file was unchanged on one
        side and deleted on the other side. The merge "passively" propagated
        that deletion, but didn't "actively" remove the file. In this case the
        file is *not* included in the `removed` set.

        d) If a file exists in only one parent and at least one of the common
        ancestors using a different filenode, then the file was changed on one
        side and removed on the other side. The merge process "actively"
        decided to drop the new change and delete the file. Unlike in the
        previous case, (c), the file included in the `removed` set.

        Summary table for merge:

        case | exists in parents | exists in gca || removed
         (a) |       both        |     *         ||   yes
         (b) |       one         |     none      ||   yes
         (c) |       one         | same filenode ||   no
         (d) |       one         |  new filenode ||   yes
        """
        return frozenset(self._removed)

    def mark_removed(self, filename):
        if 'removed' in vars(self):
            del self.removed
        self._removed.add(filename)
        self.mark_touched(filename)

    def update_removed(self, filenames):
        for f in filenames:
            self.mark_removed(f)

    @util.propertycache
    def salvaged(self):
        """files that might have been deleted by a merge, but still exists.

        During a merge, the manifest merging might select some files for
        removal, or for a removed/changed conflict. If at commit time the file
        still exists, its removal was "reverted" and the file is "salvaged"
        """
        return frozenset(self._salvaged)

    def mark_salvaged(self, filename):
        if "salvaged" in vars(self):
            del self.salvaged
        self._salvaged.add(filename)
        self.mark_touched(filename)

    def update_salvaged(self, filenames):
        for f in filenames:
            self.mark_salvaged(f)

    @util.propertycache
    def touched(self):
        """files either actively modified, added or removed"""
        return frozenset(self._touched)

    def mark_touched(self, filename):
        if 'touched' in vars(self):
            del self.touched
        self._touched.add(filename)

    def update_touched(self, filenames):
        for f in filenames:
            self.mark_touched(f)

    @util.propertycache
    def copied_from_p1(self):
        return self._p1_copies.copy()

    def mark_copied_from_p1(self, source, dest):
        if 'copied_from_p1' in vars(self):
            del self.copied_from_p1
        self._p1_copies[dest] = source

    def update_copies_from_p1(self, copies):
        for dest, source in copies.items():
            self.mark_copied_from_p1(source, dest)

    @util.propertycache
    def copied_from_p2(self):
        return self._p2_copies.copy()

    def mark_copied_from_p2(self, source, dest):
        if 'copied_from_p2' in vars(self):
            del self.copied_from_p2
        self._p2_copies[dest] = source

    def update_copies_from_p2(self, copies):
        for dest, source in copies.items():
            self.mark_copied_from_p2(source, dest)


def compute_all_files_changes(ctx):
    """compute the files changed by a revision"""
    p1 = ctx.p1()
    p2 = ctx.p2()
    if p1.rev() == nullrev and p2.rev() == nullrev:
        return _process_root(ctx)
    elif p1.rev() != nullrev and p2.rev() == nullrev:
        return _process_linear(p1, ctx)
    elif p1.rev() == nullrev and p2.rev() != nullrev:
        # In the wild, one can encounter changeset where p1 is null but p2 is not
        return _process_linear(p1, ctx, parent=2)
    elif p1.rev() == p2.rev():
        # In the wild, one can encounter such "non-merge"
        return _process_linear(p1, ctx)
    else:
        return _process_merge(p1, p2, ctx)


def _process_root(ctx):
    """compute the appropriate changed files for a changeset with no parents"""
    # Simple, there was nothing before it, so everything is added.
    md = ChangingFiles()
    manifest = ctx.manifest()
    for filename in manifest:
        md.mark_added(filename)
    return md


def _process_linear(parent_ctx, children_ctx, parent=1):
    """compute the appropriate changed files for a changeset with a single parent"""
    md = ChangingFiles()
    parent_manifest = parent_ctx.manifest()
    children_manifest = children_ctx.manifest()

    copies_candidate = []

    for filename, d in parent_manifest.diff(children_manifest).items():
        if d[1][0] is None:
            # no filenode for the "new" value, file is absent
            md.mark_removed(filename)
        else:
            copies_candidate.append(filename)
            if d[0][0] is None:
                # not filenode for the "old" value file was absent
                md.mark_added(filename)
            else:
                # filenode for both "old" and "new"
                md.mark_touched(filename)

    if parent == 1:
        copied = md.mark_copied_from_p1
    elif parent == 2:
        copied = md.mark_copied_from_p2
    else:
        assert False, "bad parent value %d" % parent

    for filename in copies_candidate:
        copy_info = children_ctx[filename].renamed()
        if copy_info:
            source, srcnode = copy_info
            copied(source, filename)

    return md


def _process_merge(p1_ctx, p2_ctx, ctx):
    """compute the appropriate changed files for a changeset with two parents

    This is a more advance case. The information we need to record is summarise
    in the following table:

    ┌──────────────┬──────────────┬──────────────┬──────────────┬──────────────┐
    │ diff ╲  diff │       ø      │ (Some, None) │ (None, Some) │ (Some, Some) │
    │  p2   ╲  p1  │              │              │              │              │
    ├──────────────┼──────────────┼──────────────┼──────────────┼──────────────┤
    │              │              │🄱  No Changes │🄳  No Changes │              │
    │  ø           │🄰  No Changes │      OR      │     OR       │🄵  No Changes │
    │              │              │🄲  Deleted[1] │🄴  Salvaged[2]│     [3]      │
    ├──────────────┼──────────────┼──────────────┼──────────────┼──────────────┤
    │              │🄶  No Changes │              │              │              │
    │ (Some, None) │      OR      │🄻  Deleted    │       ø      │      ø       │
    │              │🄷  Deleted[1] │              │              │              │
    ├──────────────┼──────────────┼──────────────┼──────────────┼──────────────┤
    │              │🄸  No Changes │              │              │              │
    │ (None, Some) │     OR       │      ø       │🄼   Added     │🄽   Merged    │
    │              │🄹  Salvaged[2]│              │   (copied?)  │   (copied?)  │
    ├──────────────┼──────────────┼──────────────┼──────────────┼──────────────┤
    │              │              │              │              │              │
    │ (Some, Some) │🄺  No Changes │      ø       │🄾   Merged    │🄿   Merged    │
    │              │     [3]      │              │   (copied?)  │   (copied?)  │
    └──────────────┴──────────────┴──────────────┴──────────────┴──────────────┘

    Special case [1]:

      The situation is:
        - parent-A:     file exists,
        - parent-B:     no file,
        - working-copy: no file.

      Detecting a "deletion" will depend on the presence of actual change on
      the "parent-A" branch:

      Subcase 🄱 or 🄶 : if the state of the file in "parent-A" is unchanged
      compared to the merge ancestors, then parent-A branch left the file
      untouched while parent-B deleted it. We simply apply the change from
      "parent-B" branch the file was automatically dropped.
      The result is:
          - file is not recorded as touched by the merge.

      Subcase 🄲 or 🄷 : otherwise, the change from parent-A branch were explicitly dropped and
      the file was "deleted again". From a user perspective, the message
      about "locally changed" while "remotely deleted" (or the other way
      around) was issued and the user chose to deleted the file.
      The result:
          - file is recorded as touched by the merge.


    Special case [2]:

      The situation is:
        - parent-A:     no file,
        - parent-B:     file,
        - working-copy: file (same content as parent-B).

      There are three subcases depending on the ancestors contents:

      - A) the file is missing in all ancestors,
      - B) at least one ancestor has the file with filenode ≠ from parent-B,
      - C) all ancestors use the same filenode as parent-B,

      Subcase (A) is the simpler, nothing happend on parent-A side while
      parent-B added it.

        The result:
            - the file is not marked as touched by the merge.

      Subcase (B) is the counter part of "Special case [1]", the file was
        modified on parent-B side, while parent-A side deleted it. However this
        time, the conflict was solved by keeping the file (and its
        modification). We consider the file as "salvaged".

        The result:
            - the file is marked as "salvaged" by the merge.

      Subcase (C) is subtle variation of the case above. In this case, the
        file in unchanged on the parent-B side and actively removed on the
        parent-A side. So the merge machinery correctly decide it should be
        removed. However, the file was explicitly restored to its parent-B
        content before the merge was commited. The file is be marked
        as salvaged too. From the merge result perspective, this is similar to
        Subcase (B), however from the merge resolution perspective they differ
        since in (C), there was some conflict not obvious solution to the
        merge (That got reversed)

    Special case [3]:

      The situation is:
        - parent-A:     file,
        - parent-B:     file (different filenode as parent-A),
        - working-copy: file (same filenode as parent-B).

      This case is in theory much simple, for this to happens, this mean the
      filenode in parent-A is purely replacing the one in parent-B (either a
      descendant, or a full new file history, see changeset). So the merge
      introduce no changes, and the file is not affected by the merge...

      However, in the wild it is possible to find commit with the above is not
      True. For example repository have some commit where the *new* node is an
      ancestor of the node in parent-A, or where parent-A and parent-B are two
      branches of the same file history, yet not merge-filenode were created
      (while the "merge" should have led to a "modification").

      Detecting such cases (and not recording the file as modified) would be a
      nice bonus. However do not any of this yet.
    """

    md = ChangingFiles()

    m = ctx.manifest()
    p1m = p1_ctx.manifest()
    p2m = p2_ctx.manifest()
    diff_p1 = p1m.diff(m)
    diff_p2 = p2m.diff(m)

    cahs = ctx.repo().changelog.commonancestorsheads(
        p1_ctx.node(), p2_ctx.node()
    )
    if not cahs:
        cahs = [nullrev]
    mas = [ctx.repo()[r].manifest() for r in cahs]

    copy_candidates = []

    # Dealing with case 🄰 happens automatically.  Since there are no entry in
    # d1 nor d2, we won't iterate on it ever.

    # Iteration over d1 content will deal with all cases, but the one in the
    # first column of the table.
    for filename, d1 in diff_p1.items():

        d2 = diff_p2.pop(filename, None)

        if d2 is None:
            # this deal with the first line of the table.
            _process_other_unchanged(md, mas, filename, d1)
        else:

            if d1[0][0] is None and d2[0][0] is None:
                # case 🄼 — both deleted the file.
                md.mark_added(filename)
                copy_candidates.append(filename)
            elif d1[1][0] is None and d2[1][0] is None:
                # case 🄻 — both deleted the file.
                md.mark_removed(filename)
            elif d1[1][0] is not None and d2[1][0] is not None:
                # case 🄽 🄾 🄿
                md.mark_merged(filename)
                copy_candidates.append(filename)
            else:
                # Impossible case, the post-merge file status cannot be None on
                # one side and Something on the other side.
                assert False, "unreachable"

    # Iteration over remaining d2 content deal with the first column of the
    # table.
    for filename, d2 in diff_p2.items():
        _process_other_unchanged(md, mas, filename, d2)

    for filename in copy_candidates:
        copy_info = ctx[filename].renamed()
        if copy_info:
            source, srcnode = copy_info
            if source in p1_ctx and p1_ctx[source].filenode() == srcnode:
                md.mark_copied_from_p1(source, filename)
            elif source in p2_ctx and p2_ctx[source].filenode() == srcnode:
                md.mark_copied_from_p2(source, filename)
    return md


def _find(manifest, filename):
    """return the associate filenode or None"""
    if filename not in manifest:
        return None
    return manifest.find(filename)[0]


def _process_other_unchanged(md, mas, filename, diff):
    source_node = diff[0][0]
    target_node = diff[1][0]

    if source_node is not None and target_node is None:
        if any(not _find(ma, filename) == source_node for ma in mas):
            # case 🄲 of 🄷
            md.mark_removed(filename)
        # else, we have case 🄱 or 🄶 : no change need to be recorded
    elif source_node is None and target_node is not None:
        if any(_find(ma, filename) is not None for ma in mas):
            # case 🄴 or 🄹
            md.mark_salvaged(filename)
        # else, we have case 🄳 or 🄸 : simple merge without intervention
    elif source_node is not None and target_node is not None:
        # case 🄵  or 🄺 : simple merge without intervention
        #
        # In buggy case where source_node is not an ancestors of target_node.
        # There should have a been a new filenode created, recording this as
        # "modified". We do not deal with them yet.
        pass
    else:
        # An impossible case, the diff algorithm should not return entry if the
        # file is missing on both side.
        assert False, "unreachable"


def _missing_from_all_ancestors(mas, filename):
    return all(_find(ma, filename) is None for ma in mas)


def computechangesetfilesadded(ctx):
    """return the list of files added in a changeset"""
    added = []
    for f in ctx.files():
        if not any(f in p for p in ctx.parents()):
            added.append(f)
    return added


def get_removal_filter(ctx, x=None):
    """return a function to detect files "wrongly" detected as `removed`

    When a file is removed relative to p1 in a merge, this
    function determines whether the absence is due to a
    deletion from a parent, or whether the merge commit
    itself deletes the file. We decide this by doing a
    simplified three way merge of the manifest entry for
    the file. There are two ways we decide the merge
    itself didn't delete a file:
    - neither parent (nor the merge) contain the file
    - exactly one parent contains the file, and that
      parent has the same filelog entry as the merge
      ancestor (or all of them if there two). In other
      words, that parent left the file unchanged while the
      other one deleted it.
    One way to think about this is that deleting a file is
    similar to emptying it, so the list of changed files
    should be similar either way. The computation
    described above is not done directly in _filecommit
    when creating the list of changed files, however
    it does something very similar by comparing filelog
    nodes.
    """

    if x is not None:
        p1, p2, m1, m2 = x
    else:
        p1 = ctx.p1()
        p2 = ctx.p2()
        m1 = p1.manifest()
        m2 = p2.manifest()

    @util.cachefunc
    def mas():
        p1n = p1.node()
        p2n = p2.node()
        cahs = ctx.repo().changelog.commonancestorsheads(p1n, p2n)
        if not cahs:
            cahs = [nullrev]
        return [ctx.repo()[r].manifest() for r in cahs]

    def deletionfromparent(f):
        if f in m1:
            return f not in m2 and all(
                f in ma and ma.find(f) == m1.find(f) for ma in mas()
            )
        elif f in m2:
            return all(f in ma and ma.find(f) == m2.find(f) for ma in mas())
        else:
            return True

    return deletionfromparent


def computechangesetfilesremoved(ctx):
    """return the list of files removed in a changeset"""
    removed = []
    for f in ctx.files():
        if f not in ctx:
            removed.append(f)
    if removed:
        rf = get_removal_filter(ctx)
        removed = [r for r in removed if not rf(r)]
    return removed


def computechangesetfilesmerged(ctx):
    """return the list of files merged in a changeset"""
    merged = []
    if len(ctx.parents()) < 2:
        return merged
    for f in ctx.files():
        if f in ctx:
            fctx = ctx[f]
            parents = fctx._filelog.parents(fctx._filenode)
            if parents[1] != nullid:
                merged.append(f)
    return merged


def computechangesetcopies(ctx):
    """return the copies data for a changeset

    The copies data are returned as a pair of dictionnary (p1copies, p2copies).

    Each dictionnary are in the form: `{newname: oldname}`
    """
    p1copies = {}
    p2copies = {}
    p1 = ctx.p1()
    p2 = ctx.p2()
    narrowmatch = ctx._repo.narrowmatch()
    for dst in ctx.files():
        if not narrowmatch(dst) or dst not in ctx:
            continue
        copied = ctx[dst].renamed()
        if not copied:
            continue
        src, srcnode = copied
        if src in p1 and p1[src].filenode() == srcnode:
            p1copies[dst] = src
        elif src in p2 and p2[src].filenode() == srcnode:
            p2copies[dst] = src
    return p1copies, p2copies


def encodecopies(files, copies):
    items = []
    for i, dst in enumerate(files):
        if dst in copies:
            items.append(b'%d\0%s' % (i, copies[dst]))
    if len(items) != len(copies):
        raise error.ProgrammingError(
            b'some copy targets missing from file list'
        )
    return b"\n".join(items)


def decodecopies(files, data):
    try:
        copies = {}
        if not data:
            return copies
        for l in data.split(b'\n'):
            strindex, src = l.split(b'\0')
            i = int(strindex)
            dst = files[i]
            copies[dst] = src
        return copies
    except (ValueError, IndexError):
        # Perhaps someone had chosen the same key name (e.g. "p1copies") and
        # used different syntax for the value.
        return None


def encodefileindices(files, subset):
    subset = set(subset)
    indices = []
    for i, f in enumerate(files):
        if f in subset:
            indices.append(b'%d' % i)
    return b'\n'.join(indices)


def decodefileindices(files, data):
    try:
        subset = []
        if not data:
            return subset
        for strindex in data.split(b'\n'):
            i = int(strindex)
            if i < 0 or i >= len(files):
                return None
            subset.append(files[i])
        return subset
    except (ValueError, IndexError):
        # Perhaps someone had chosen the same key name (e.g. "added") and
        # used different syntax for the value.
        return None


# see mercurial/helptext/internals/revlogs.txt for details about the format

ACTION_MASK = int("111" "00", 2)
# note: untouched file used as copy source will as `000` for this mask.
ADDED_FLAG = int("001" "00", 2)
MERGED_FLAG = int("010" "00", 2)
REMOVED_FLAG = int("011" "00", 2)
SALVAGED_FLAG = int("100" "00", 2)
TOUCHED_FLAG = int("101" "00", 2)

COPIED_MASK = int("11", 2)
COPIED_FROM_P1_FLAG = int("10", 2)
COPIED_FROM_P2_FLAG = int("11", 2)

# structure is <flag><filename-end><copy-source>
INDEX_HEADER = struct.Struct(">L")
INDEX_ENTRY = struct.Struct(">bLL")


def encode_files_sidedata(files):
    all_files = set(files.touched)
    all_files.update(files.copied_from_p1.values())
    all_files.update(files.copied_from_p2.values())
    all_files = sorted(all_files)
    file_idx = {f: i for (i, f) in enumerate(all_files)}
    file_idx[None] = 0

    chunks = [INDEX_HEADER.pack(len(all_files))]

    filename_length = 0
    for f in all_files:
        filename_size = len(f)
        filename_length += filename_size
        flag = 0
        if f in files.added:
            flag |= ADDED_FLAG
        elif f in files.merged:
            flag |= MERGED_FLAG
        elif f in files.removed:
            flag |= REMOVED_FLAG
        elif f in files.salvaged:
            flag |= SALVAGED_FLAG
        elif f in files.touched:
            flag |= TOUCHED_FLAG

        copy = None
        if f in files.copied_from_p1:
            flag |= COPIED_FROM_P1_FLAG
            copy = files.copied_from_p1.get(f)
        elif f in files.copied_from_p2:
            copy = files.copied_from_p2.get(f)
            flag |= COPIED_FROM_P2_FLAG
        copy_idx = file_idx[copy]
        chunks.append(INDEX_ENTRY.pack(flag, filename_length, copy_idx))
    chunks.extend(all_files)
    return {sidedatamod.SD_FILES: b''.join(chunks)}


def decode_files_sidedata(sidedata):
    md = ChangingFiles()
    raw = sidedata.get(sidedatamod.SD_FILES)

    if raw is None:
        return md

    copies = []
    all_files = []

    assert len(raw) >= INDEX_HEADER.size
    total_files = INDEX_HEADER.unpack_from(raw, 0)[0]

    offset = INDEX_HEADER.size
    file_offset_base = offset + (INDEX_ENTRY.size * total_files)
    file_offset_last = file_offset_base

    assert len(raw) >= file_offset_base

    for idx in range(total_files):
        flag, file_end, copy_idx = INDEX_ENTRY.unpack_from(raw, offset)
        file_end += file_offset_base
        filename = raw[file_offset_last:file_end]
        filesize = file_end - file_offset_last
        assert len(filename) == filesize
        offset += INDEX_ENTRY.size
        file_offset_last = file_end
        all_files.append(filename)
        if flag & ACTION_MASK == ADDED_FLAG:
            md.mark_added(filename)
        elif flag & ACTION_MASK == MERGED_FLAG:
            md.mark_merged(filename)
        elif flag & ACTION_MASK == REMOVED_FLAG:
            md.mark_removed(filename)
        elif flag & ACTION_MASK == SALVAGED_FLAG:
            md.mark_salvaged(filename)
        elif flag & ACTION_MASK == TOUCHED_FLAG:
            md.mark_touched(filename)

        copied = None
        if flag & COPIED_MASK == COPIED_FROM_P1_FLAG:
            copied = md.mark_copied_from_p1
        elif flag & COPIED_MASK == COPIED_FROM_P2_FLAG:
            copied = md.mark_copied_from_p2

        if copied is not None:
            copies.append((copied, filename, copy_idx))

    for copied, filename, copy_idx in copies:
        copied(all_files[copy_idx], filename)

    return md


def _getsidedata(srcrepo, rev):
    ctx = srcrepo[rev]
    files = compute_all_files_changes(ctx)
    return encode_files_sidedata(files), files.has_copies_info


def getsidedataadder(srcrepo, destrepo):
    use_w = srcrepo.ui.configbool(b'experimental', b'worker.repository-upgrade')
    if pycompat.iswindows or not use_w:
        return _get_simple_sidedata_adder(srcrepo, destrepo)
    else:
        return _get_worker_sidedata_adder(srcrepo, destrepo)


def _sidedata_worker(srcrepo, revs_queue, sidedata_queue, tokens):
    """The function used by worker precomputing sidedata

    It read an input queue containing revision numbers
    It write in an output queue containing (rev, <sidedata-map>)

    The `None` input value is used as a stop signal.

    The `tokens` semaphore is user to avoid having too many unprocessed
    entries. The workers needs to acquire one token before fetching a task.
    They will be released by the consumer of the produced data.
    """
    tokens.acquire()
    rev = revs_queue.get()
    while rev is not None:
        data = _getsidedata(srcrepo, rev)
        sidedata_queue.put((rev, data))
        tokens.acquire()
        rev = revs_queue.get()
    # processing of `None` is completed, release the token.
    tokens.release()


BUFF_PER_WORKER = 50


def _get_worker_sidedata_adder(srcrepo, destrepo):
    """The parallel version of the sidedata computation

    This code spawn a pool of worker that precompute a buffer of sidedata
    before we actually need them"""
    # avoid circular import copies -> scmutil -> worker -> copies
    from . import worker

    nbworkers = worker._numworkers(srcrepo.ui)

    tokens = multiprocessing.BoundedSemaphore(nbworkers * BUFF_PER_WORKER)
    revsq = multiprocessing.Queue()
    sidedataq = multiprocessing.Queue()

    assert srcrepo.filtername is None
    # queue all tasks beforehand, revision numbers are small and it make
    # synchronisation simpler
    #
    # Since the computation for each node can be quite expensive, the overhead
    # of using a single queue is not revelant. In practice, most computation
    # are fast but some are very expensive and dominate all the other smaller
    # cost.
    for r in srcrepo.changelog.revs():
        revsq.put(r)
    # queue the "no more tasks" markers
    for i in range(nbworkers):
        revsq.put(None)

    allworkers = []
    for i in range(nbworkers):
        args = (srcrepo, revsq, sidedataq, tokens)
        w = multiprocessing.Process(target=_sidedata_worker, args=args)
        allworkers.append(w)
        w.start()

    # dictionnary to store results for revision higher than we one we are
    # looking for. For example, if we need the sidedatamap for 42, and 43 is
    # received, when shelve 43 for later use.
    staging = {}

    def sidedata_companion(revlog, rev):
        data = {}, False
        if util.safehasattr(revlog, b'filteredrevs'):  # this is a changelog
            # Is the data previously shelved ?
            sidedata = staging.pop(rev, None)
            if sidedata is None:
                # look at the queued result until we find the one we are lookig
                # for (shelve the other ones)
                r, data = sidedataq.get()
                while r != rev:
                    staging[r] = data
                    r, sidedata = sidedataq.get()
            tokens.release()
        sidedata, has_copies_info = data
        new_flag = 0
        if has_copies_info:
            new_flag = sidedataflag.REVIDX_HASCOPIESINFO
        return False, (), sidedata, new_flag, 0

    return sidedata_companion


def _get_simple_sidedata_adder(srcrepo, destrepo):
    """The simple version of the sidedata computation

    It just compute it in the same thread on request"""

    def sidedatacompanion(revlog, rev):
        sidedata, has_copies_info = {}, False
        if util.safehasattr(revlog, 'filteredrevs'):  # this is a changelog
            sidedata, has_copies_info = _getsidedata(srcrepo, rev)
        new_flag = 0
        if has_copies_info:
            new_flag = sidedataflag.REVIDX_HASCOPIESINFO

        return False, (), sidedata, new_flag, 0

    return sidedatacompanion


def getsidedataremover(srcrepo, destrepo):
    def sidedatacompanion(revlog, rev):
        f = ()
        if util.safehasattr(revlog, 'filteredrevs'):  # this is a changelog
            if revlog.flags(rev) & sidedataflag.REVIDX_SIDEDATA:
                f = (
                    sidedatamod.SD_P1COPIES,
                    sidedatamod.SD_P2COPIES,
                    sidedatamod.SD_FILESADDED,
                    sidedatamod.SD_FILESREMOVED,
                )
        return False, f, {}, 0, sidedataflag.REVIDX_HASCOPIESINFO

    return sidedatacompanion