Routino SVN Repository Browser

---

Revision 1608 - (view) (download) (as text) (annotate) - [select for diffs]
Modified Sat Oct 18 12:50:40 2014 UTC (10 years, 4 months ago) by amb
File length: 26026 byte(s)
Diff to previous 1607

Use the waysx->idata array when logging duplicate segments rather than looking
up the wayx and using its id, also saves mapping the ways into memory.

---

Revision 1607 - (view) (download) (as text) (annotate) - [select for diffs]
Modified Sat Oct 18 12:41:54 2014 UTC (10 years, 4 months ago) by amb
File length: 26238 byte(s)
Diff to previous 1605

Use the nodesx->idata array when logging duplicate segments rather than looking
up the nodex and using its id.

---

Revision 1605 - (view) (download) (as text) (annotate) - [select for diffs]
Modified Sat Oct 18 10:30:57 2014 UTC (10 years, 4 months ago) by amb
File length: 26203 byte(s)
Diff to previous 1603

Free memory that it allocated by IndexSegments() when no longer needed rather
than holding on to it.

---

Revision 1603 - (view) (download) (as text) (annotate) - [select for diffs]
Modified Mon Oct 13 18:25:27 2014 UTC (10 years, 5 months ago) by amb
File length: 26308 byte(s)
Diff to previous 1598

Try to speed up the search for the closest node/segment by minimising the number
of nodes that are examined in detail.

---

Revision 1598 - (view) (download) (as text) (annotate) - [select for diffs]
Modified Sat Sep 27 16:35:09 2014 UTC (10 years, 5 months ago) by amb
File length: 26310 byte(s)
Diff to previous 1593

Add a '--logmemory' option to planetsplitter which will report the maximum
memory in use (allocated and mapped files) during each step of the processing.

---

Revision 1593 - (view) (download) (as text) (annotate) - [select for diffs]
Modified Fri Sep 26 14:48:01 2014 UTC (10 years, 5 months ago) by amb
File length: 25767 byte(s)
Diff to previous 1448

Free segmentsx->usedway in FreeSegmentList() rather than letting it leak.

---

Revision 1448 - (view) (download) (as text) (annotate) - [select for diffs]
Modified Tue Jul 2 17:52:01 2013 UTC (11 years, 8 months ago) by amb
File length: 25712 byte(s)
Diff to previous 1432

Revert r1432 because even though it seemed like it should have been faster in
slim mode it was actually much slower on the routino.org virtual server.

---

Revision 1432 - (view) (download) (as text) (annotate) - [select for diffs]
Modified Fri Jun 28 14:41:56 2013 UTC (11 years, 8 months ago) by amb
File length: 26038 byte(s)
Diff to previous 1431

Keep the next2 pointer in memory while pruning rather than in the segmentx
object.

---

Revision 1431 - (view) (download) (as text) (annotate) - [select for diffs]
Modified Fri Jun 28 14:39:14 2013 UTC (11 years, 8 months ago) by amb
File length: 25712 byte(s)
Diff to previous 1430

Revert the last three changes because r1430 didn't work and r1428+r1429 didn't
give any speed advantage and was possibly marginally slower.

---

Revision 1430 - (view) (download) (as text) (annotate) - [select for diffs]
Modified Fri Jun 28 14:07:07 2013 UTC (11 years, 8 months ago) by amb
File length: 24595 byte(s)
Diff to previous 1429

The IndexSegments() function can now process the file in forward order rather
than reverse.

---

Revision 1429 - (view) (download) (as text) (annotate) - [select for diffs]
Modified Thu Jun 27 19:02:48 2013 UTC (11 years, 8 months ago) by amb
File length: 25168 byte(s)
Diff to previous 1428

Fixed error with last checkin.

---

Revision 1428 - (view) (download) (as text) (annotate) - [select for diffs]
Modified Thu Jun 27 18:29:33 2013 UTC (11 years, 8 months ago) by amb
File length: 25109 byte(s)
Diff to previous 1427

Put the next1 pointer in the segmentx object rather than in-memory.

---

Revision 1427 - (view) (download) (as text) (annotate) - [select for diffs]
Modified Thu Jun 27 18:01:14 2013 UTC (11 years, 8 months ago) by amb
File length: 25712 byte(s)
Diff to previous 1426

Revert the last change because, paradoxically, it was faster to create the
database (as expected) but slower to route.

---

Revision 1426 - (view) (download) (as text) (annotate) - [select for diffs]
Modified Wed Jun 26 18:26:27 2013 UTC (11 years, 8 months ago) by amb
File length: 23597 byte(s)
Diff to previous 1422

Sort the nodes geographically at the beginning rather than at the end.

---

Revision 1422 - (view) (download) (as text) (annotate) - [select for diffs]
Modified Mon Jun 24 18:53:09 2013 UTC (11 years, 8 months ago) by amb
File length: 25712 byte(s)
Diff to previous 1414

Change the IndexSegments() function so that in slim mode larger chunks are read
from the file (since it reads sequentially but in reverse through the file).

---

Revision 1414 - (view) (download) (as text) (annotate) - [select for diffs]
Modified Sat Jun 22 18:49:33 2013 UTC (11 years, 8 months ago) by amb
File length: 25159 byte(s)
Diff to previous 1410

Use SlimMapFile() and SlimUnmapFile() as the function names for the slim mode of
operation to open and close the files (to hide the fact that they are files
being opened for reading or writing unbuffered).

---

Revision 1410 - (view) (download) (as text) (annotate) - [select for diffs]
Modified Fri Jun 21 15:48:19 2013 UTC (11 years, 8 months ago) by amb
File length: 25219 byte(s)
Diff to previous 1409

Rename the functions for unbuffered file access to make this clear.

---

Revision 1409 - (view) (download) (as text) (annotate) - [select for diffs]
Modified Fri Jun 21 14:53:01 2013 UTC (11 years, 8 months ago) by amb
File length: 25139 byte(s)
Diff to previous 1408

Use the new buffered functions in the filesort functions.

---

Revision 1408 - (view) (download) (as text) (annotate) - [select for diffs]
Modified Fri Jun 21 14:43:37 2013 UTC (11 years, 8 months ago) by amb
File length: 25011 byte(s)
Diff to previous 1402

Use the new functions for buffering while reading when looping through files
other than the ones already done that use the FILESORT_VARINT method of storing
data.

---

Revision 1402 - (view) (download) (as text) (annotate) - [select for diffs]
Modified Wed Jun 19 18:58:05 2013 UTC (11 years, 8 months ago) by amb
File length: 24891 byte(s)
Diff to previous 1355

Use the buffered write functions when creating the nodes, segments, ways and
relation raw files.

---

Revision 1355 - (view) (download) (as text) (annotate) - [select for diffs]
Modified Fri May 31 14:06:50 2013 UTC (11 years, 9 months ago) by amb
File length: 24867 byte(s)
Diff to previous 1354

Small changes to comments and log messages.

---

Revision 1354 - (view) (download) (as text) (annotate) - [select for diffs]
Modified Thu May 30 18:16:25 2013 UTC (11 years, 9 months ago) by amb
File length: 24875 byte(s)
Diff to previous 1351

Don't de-duplicate when sorting segments (now done in ProcessSegments()
function).

---

Revision 1351 - (view) (download) (as text) (annotate) - [select for diffs]
Modified Thu May 30 17:33:21 2013 UTC (11 years, 9 months ago) by amb
File length: 25867 byte(s)
Diff to previous 1350

Merge the RemoveBadSegments() and MeasureSegments() functions.  Saves one
read/write iteration through the segments file.

---

Revision 1350 - (view) (download) (as text) (annotate) - [select for diffs]
Modified Thu May 30 16:53:27 2013 UTC (11 years, 9 months ago) by amb
File length: 27535 byte(s)
Diff to previous 1349

Delete the non-highway nodes by searching for them in the ways rather than
marking them when processing the segments.

---

Revision 1349 - (view) (download) (as text) (annotate) - [select for diffs]
Added Wed May 29 17:33:26 2013 UTC (11 years, 9 months ago) by amb
File length: 28105 byte(s)
Diff to previous 1347

Simplify the segments by using the node and way index instead of node and way id
which avoids lots of IndexNodeX() lookups.  Move some other code around to cope
with it.