static void parseFixedStepSection(struct lineFile *lf, boolean clipDontDie, struct lm *lm, int itemsPerSlot, char *chrom, bits32 chromSize, bits32 span, bits32 sectionStart, bits32 step, struct bwgSection **pSectionList) /* Read the single column data in section until get to end. */ { struct lm *lmLocal = lmInit(0); /* Stream through section until get to end of file or next section, * adding values from single column to list. */ char *words[1]; char *line; struct bwgFixedStepItem *item, *itemList = NULL; int originalSectionSize = 0; bits32 sectionEnd = sectionStart; while (lineFileNextReal(lf, &line)) { if (steppedSectionEnd(line, 1)) { lineFileReuse(lf); break; } chopLine(line, words); lmAllocVar(lmLocal, item); item->val = lineFileNeedDouble(lf, words, 0); if (sectionEnd + span > chromSize) { warn("line %d of %s: chromosome %s has %u bases, but item ends at %u", lf->lineIx, lf->fileName, chrom, chromSize, sectionEnd + span); if (!clipDontDie) noWarnAbort(); } else { slAddHead(&itemList, item); ++originalSectionSize; } sectionEnd += step; } slReverse(&itemList); /* Break up into sections of no more than items-per-slot size, and convert to packed format. */ int sizeLeft = originalSectionSize; for (item = itemList; item != NULL; ) { /* Figure out size of this section */ int sectionSize = sizeLeft; if (sectionSize > itemsPerSlot) sectionSize = itemsPerSlot; sizeLeft -= sectionSize; /* Allocate and fill in section. */ struct bwgSection *section; lmAllocVar(lm, section); section->chrom = chrom; section->start = sectionStart; sectionStart += sectionSize * step; section->end = sectionStart - step + span; section->type = bwgTypeFixedStep; section->itemStep = step; section->itemSpan = span; section->itemCount = sectionSize; /* Allocate array for data, and copy from list to array representation */ struct bwgFixedStepPacked *packed; /* An array */ section->items.fixedStepPacked = lmAllocArray(lm, packed, sectionSize); int i; for (i=0; i<sectionSize; ++i) { packed->val = item->val; item = item->next; ++packed; } /* Add section to list. */ slAddHead(pSectionList, section); } lmCleanup(&lmLocal); }
static void parseVariableStepSection(struct lineFile *lf, boolean clipDontDie, struct lm *lm, int itemsPerSlot, char *chrom, int chromSize, bits32 span, struct bwgSection **pSectionList) /* Read the single column data in section until get to end. */ { struct lm *lmLocal = lmInit(0); /* Stream through section until get to end of file or next section, * adding values from single column to list. */ char *words[2]; char *line; struct bwgVariableStepItem *item, *nextItem, *itemList = NULL; int originalSectionSize = 0; while (lineFileNextReal(lf, &line)) { if (steppedSectionEnd(line, 2)) { lineFileReuse(lf); break; } chopLine(line, words); lmAllocVar(lmLocal, item); int start = lineFileNeedNum(lf, words, 0); if (start <= 0) { errAbort("line %d of %s: zero or negative chromosome coordinate not allowed", lf->lineIx, lf->fileName); } item->start = start - 1; item->val = lineFileNeedDouble(lf, words, 1); if (item->start + span > chromSize) { warn("line %d of %s: chromosome %s has %u bases, but item ends at %u", lf->lineIx, lf->fileName, chrom, chromSize, item->start + span); if (!clipDontDie) noWarnAbort(); } else { slAddHead(&itemList, item); ++originalSectionSize; } } slSort(&itemList, bwgVariableStepItemCmp); /* Make sure no overlap between items. */ if (itemList != NULL) { item = itemList; for (nextItem = item->next; nextItem != NULL; nextItem = nextItem->next) { if (item->start + span > nextItem->start) errAbort("Overlap on %s between items starting at %d and %d.\n" "Please remove overlaps and try again", chrom, item->start, nextItem->start); item = nextItem; } } /* Break up into sections of no more than items-per-slot size. */ int sizeLeft = originalSectionSize; for (item = itemList; item != NULL; ) { /* Figure out size of this section */ int sectionSize = sizeLeft; if (sectionSize > itemsPerSlot) sectionSize = itemsPerSlot; sizeLeft -= sectionSize; /* Convert from list to array representation. */ struct bwgVariableStepPacked *packed, *p; p = lmAllocArray(lm, packed, sectionSize); int i; for (i=0; i<sectionSize; ++i) { p->start = item->start; p->val = item->val; item = item->next; ++p; } /* Fill in section and add it to list. */ struct bwgSection *section; lmAllocVar(lm, section); section->chrom = chrom; section->start = packed[0].start; section->end = packed[sectionSize-1].start + span; section->type = bwgTypeVariableStep; section->items.variableStepPacked = packed; section->itemSpan = span; section->itemCount = sectionSize; slAddHead(pSectionList, section); } lmCleanup(&lmLocal); }
static void parseSteppedSection(struct lineFile *lf, char *initialLine, struct pos *pos) /* Parse out stepped section, adding info about it to pos. */ { /* Parse out first word of initial line and make sure it is something we recognize. */ char *typeWord = nextWord(&initialLine); enum bwgSectionType type = bwgTypeFixedStep; if (sameString(typeWord, "variableStep")) type = bwgTypeVariableStep; else if (sameString(typeWord, "fixedStep")) type = bwgTypeFixedStep; else errAbort("Unknown type %s\n", typeWord); /* Set up defaults for values we hope to parse out of rest of line. */ bits32 start = 0; char *chrom = NULL; /* Parse out var=val pairs. */ char *varEqVal; while ((varEqVal = nextWord(&initialLine)) != NULL) { char *wordPairs[2]; int wc = chopByChar(varEqVal, '=', wordPairs, 2); if (wc != 2) errAbort("strange var=val pair line %d of %s", lf->lineIx, lf->fileName); char *var = wordPairs[0]; char *val = wordPairs[1]; if (sameString(var, "chrom")) chrom = cloneString(val); else if (sameString(var, "start")) start = parseUnsignedVal(lf, var, val); } /* Check that we have all that are required and no more, and call type-specific routine to parse * rest of section. */ if (chrom == NULL) errAbort("Missing chrom= setting line %d of %s\n", lf->lineIx, lf->fileName); pos->chrom = chrom; if (type == bwgTypeFixedStep) { if (start == 0) errAbort("Missing start= setting line %d of %s\n", lf->lineIx, lf->fileName); for (;;) { char *line; if (!lineFileNextReal(lf, &line)) break; if (steppedSectionEnd(line, 1)) { lineFileReuse(lf); break; } } } else { for (;;) { char *line; if (!lineFileNextReal(lf, &line)) break; if (steppedSectionEnd(line, 2)) { lineFileReuse(lf); break; } bits32 s = atoi(line); if (start == 0 || s < start) start = s; } } pos->start = start-1; }