lib/dtdParse.c File Reference

#include "common.h"
#include "hash.h"
#include "linefile.h"
#include "dystring.h"
#include "obscure.h"
#include "dtdParse.h"

Include dependency graph for dtdParse.c:

Go to the source code of this file.

Functions
static void	syntaxError (struct lineFile *lf)
static char *	needNextWord (char **pLine, struct lineFile *lf)
void	needQuotedString (char *in, char *out, struct lineFile *lf, char **retNext)
static boolean	isAllUpper (char *s)
static boolean	isAllLower (char *s)
static char *	mixedCaseName (char *prefix, char *orig)
static struct hash *	initialEntityHash ()
static struct dtdElement *	parseElement (char *prefix, char *textField, char *line, struct hash *elHash, struct lineFile *lf)
static void	parseAttribute (char *line, char *textField, struct hash *elHash, struct lineFile *lf)
void	parseEntity (struct hash *entityHash, struct hash *predefEntityHash, char *line, struct lineFile *lf)
static void	fixupChildRefs (struct dtdElement *elList, struct hash *elHash, char *fileName)
static char *	eatComment (struct lineFile *lf, char *line)
static void	expandEntities (char *s, struct hash *entityHash, struct lineFile *lf, struct dyString *dest)
static char *	dtdxTag (struct lineFile *lf, struct hash *entityHash, struct dyString *buf)
void	dtdParse (char *fileName, char *prefix, char *textField, struct dtdElement **retList, struct hash **retHash)
void	dtdElementDump (struct dtdElement *el, FILE *f)

---

Function Documentation

void dtdElementDump	(	struct dtdElement *	el,
		FILE *	f
	)

Definition at line 439 of file dtdParse.c.

References dtdElement::attributes, dtdElement::children, dtdElChild::copyCode, dtdElChild::isOr, dtdElement::mixedCaseName, dtdAttribute::name, dtdElChild::name, dtdElement::name, dtdAttribute::next, dtdElChild::next, dtdAttribute::required, dtdElement::textType, dtdAttribute::type, and dtdAttribute::usual.

{
struct dtdElChild *ec;
struct dtdAttribute *att;
fprintf(f, "%s %s (", el->name, el->mixedCaseName);
for (ec = el->children; ec != NULL; ec = ec->next)
    {
    fprintf(f, "%s", ec->name);
    if (ec->copyCode != '1')
        fprintf(f, "%c", ec->copyCode);
    if (ec->isOr)
        fprintf(f, " (isOr)");
    if (ec->next != NULL)
        fprintf(f, ", ");
    }
fprintf(f, ")");
if (el->textType != NULL)
    fprintf(f, " (%s)", el->textType);
fprintf(f, "\n");
for (att = el->attributes; att != NULL; att = att->next)
    {
    fprintf(f, "  %s %s %s %s\n",
        att->name, att->type, (att->usual ? att->usual : "n/a"),  
        (att->required ? "required" : "optional"));
    }
}

void dtdParse	(	char *	fileName,
		char *	prefix,
		char *	textField,
		struct dtdElement **	retList,
		struct hash **	retHash
	)

Definition at line 371 of file dtdParse.c.

References dtdxTag(), dyStringFree, dyStringNew, eatComment(), errAbort(), fixupChildRefs(), freeHashAndVals(), initialEntityHash(), lineFileClose(), lineFileOpen(), needNextWord(), newHash(), parseAttribute(), parseElement(), parseEntity(), sameWord, slAddHead, slReverse(), startsWith(), syntaxError(), trimSpaces(), and TRUE.

{
struct hash *elHash = newHash(8);
struct hash *entityHash = initialEntityHash();
struct hash *predefEntityHash = initialEntityHash();
struct dtdElement *elList = NULL, *el;
struct lineFile *lf = lineFileOpen(fileName, TRUE);
char *line, *word;
struct dyString *buf = dyStringNew(0);

if (prefix == NULL)
    prefix = "";
if (textField == NULL)
    textField = "text";
while ((line = dtdxTag(lf, entityHash, buf)) != NULL)
    {
    line = trimSpaces(line);
    if (line == NULL || line[0] == 0 || line[0] == '#')
        continue;
    if (startsWith("<!--", line))
        {
        line = eatComment(lf, line);
        if (line == NULL)
            continue;
        }
    if (!startsWith("<!", line))
        syntaxError(lf);
    line += 2;
    word = needNextWord(&line, lf);
    if (sameWord("ELEMENT", word))
        {
        el = parseElement(prefix, textField, line, elHash, lf);
        slAddHead(&elList, el);
        }
    else if (sameWord("ATTLIST", word))
        {
        parseAttribute(line, textField, elHash, lf);
        }
    else if (sameWord("ENTITY", word))
        {
        parseEntity(entityHash, predefEntityHash, line, lf);
        }
    else
        {
        errAbort("Don't understand %s line %d of %s", word, lf->lineIx, lf->fileName);
        }
    }
lineFileClose(&lf);
dyStringFree(&buf);
slReverse(&elList);
fixupChildRefs(elList, elHash, fileName);
freeHashAndVals(&entityHash);
freeHashAndVals(&predefEntityHash);
*retHash = elHash;
*retList = elList;
}

Here is the call graph for this function:

static char* dtdxTag	(	struct lineFile *	lf,
		struct hash *	entityHash,
		struct dyString *	buf
	)			[static]

Definition at line 345 of file dtdParse.c.

References dyStringAppendC(), dyStringClear, errAbort(), expandEntities(), lineFile::fileName, lineFileNext(), lineFileNextReal(), lineFile::lineIx, dyString::string, dyString::stringSize, and trimSpaces().

Referenced by dtdParse().

{
char *line;

/* Skip until get a line that starts with '<' */
if (!lineFileNextReal(lf,  &line))
    return NULL;
line = trimSpaces(line);
if (line[0] != '<')
    errAbort("Text outside of a tag line %d of %s", lf->lineIx, lf->fileName);
dyStringClear(buf);
for (;;)
    {
    expandEntities(line, entityHash, lf, buf);
    if (buf->string[buf->stringSize-1] == '>')
         break;
    dyStringAppendC(buf, ' ');
    if (!lineFileNext(lf, &line, NULL))
        errAbort("End of file %s inside of a tag.", lf->fileName);
    line = trimSpaces(line);
    }
return buf->string;
}

Here is the call graph for this function:

Here is the caller graph for this function:

static char* eatComment	(	struct lineFile *	lf,
		char *	line
	)			[static]

Definition at line 298 of file dtdParse.c.

References lineFileNext(), skipLeadingSpaces(), and stringIn.

Referenced by dtdParse().

{
char *s;
for (;;)
    {
    if ((s = stringIn("-->", line)) != NULL)
        {
        line = skipLeadingSpaces(s+3);
        if (line[0] == 0)
            line = NULL;
        return line;
        }
    if (!lineFileNext(lf, &line, NULL))
        return NULL;
    }
}

Here is the call graph for this function:

Here is the caller graph for this function:

static void expandEntities	(	char *	s,
		struct hash *	entityHash,
		struct lineFile *	lf,
		struct dyString *	dest
	)			[static]

Definition at line 316 of file dtdParse.c.

References dyStringAppend(), dyStringAppendC(), errAbort(), lineFile::fileName, hashFindVal(), lineFile::lineIx, and name.

Referenced by dtdxTag().

{
char c;
while ((c = *s++) != 0)
    {
    if (c == '%' && !isspace(s[0]))
        {
        char *name = s;
        char *end = strchr(s, ';');
        char *value;
        if (end == NULL)
            errAbort("Can't find ; after %% to close entity line %d of %s",
                lf->lineIx, lf->fileName);
        *end++ = 0;
        s = end;
        value = hashFindVal(entityHash, name);
        if (value == NULL)
            errAbort("Entity %%%s; is not defined line %d of %s",
                name, lf->lineIx, lf->fileName);
        dyStringAppend(dest, value);
        }
    else
        dyStringAppendC(dest, c);
    }
}

Here is the call graph for this function:

Here is the caller graph for this function:

static void fixupChildRefs	(	struct dtdElement *	elList,
		struct hash *	elHash,
		char *	fileName
	)			[static]

Definition at line 281 of file dtdParse.c.

References dtdElement::children, dtdElChild::el, errAbort(), hashFindVal(), dtdElement::lineIx, dtdElement::name, dtdElChild::name, dtdElChild::next, and dtdElement::next.

Referenced by dtdParse().

{
struct dtdElement *el, *child;
struct dtdElChild *ec;
for (el = elList; el != NULL; el = el->next)
    {
    for (ec = el->children; ec != NULL; ec = ec->next)
        {
        if ((child = hashFindVal(elHash, ec->name)) == NULL)
            errAbort("%s's child %s undefined line %d of %s", el->name, ec->name, el->lineIx, fileName);
        ec->el = child;
        }
    }
}

Here is the call graph for this function:

Here is the caller graph for this function:

static struct hash* initialEntityHash

(

)

[static, read]

Definition at line 103 of file dtdParse.c.

References cloneString(), hashAdd(), and hashNew.

Referenced by dtdParse().

{
struct hash *hash = hashNew(0);
hashAdd(hash, "INTEGER", cloneString("#INT"));
hashAdd(hash, "REAL", cloneString("#FLOAT"));
hashAdd(hash, "INT", cloneString("INT"));
hashAdd(hash, "FLOAT", cloneString("FLOAT"));
return hash;
}

Here is the call graph for this function:

Here is the caller graph for this function:

static boolean isAllLower

(

char *

s

)

[static]

Definition at line 49 of file dtdParse.c.

References FALSE, and TRUE.

Referenced by mixedCaseName().

{
char c;
while ((c = *s++) != 0)
    {
    if (isalpha(c) && !islower(c))
        return FALSE;
    }
return TRUE;
}

Here is the caller graph for this function:

static boolean isAllUpper

(

char *

s

)

[static]

Definition at line 36 of file dtdParse.c.

References FALSE, and TRUE.

Referenced by mixedCaseName().

{
char c;
while ((c = *s++) != 0)
    {
    if (isalpha(c) && !isupper(c))
        return FALSE;
    }
return TRUE;
}

Here is the caller graph for this function:

static char* mixedCaseName	(	char *	prefix,
		char *	orig
	)			[static]

Definition at line 63 of file dtdParse.c.

References FALSE, isAllLower(), isAllUpper(), needMem(), and TRUE.

Referenced by parseAttribute(), and parseElement().

{
char *mixed;
char *d, *s = orig;
char c;
int prefixLen = strlen(prefix), len;
boolean nextUpper;
boolean allUpper = isAllUpper(orig); 
boolean allLower = isAllLower(orig);
boolean initiallyMixed = (!allUpper && !allLower);

/* Allocate string big enough for prefix and all. */
len = strlen(orig) + prefixLen;
mixed = d = needMem(len+1);
strcpy(d, prefix);
d += prefixLen;
nextUpper = (prefixLen > 0);

for (;;)
   {
   c = *s++;
   if (c == '_' || c == '-' || c == ':')
       nextUpper = TRUE;
   else
       {
       if (nextUpper)
           c = toupper(c);
       else if (!initiallyMixed)
           c = tolower(c);
       nextUpper = FALSE;
       *d++ = c;
       if (c == 0)
           break;
       }
   }
return mixed;
}

Here is the call graph for this function:

Here is the caller graph for this function:

static char* needNextWord	(	char **	pLine,
		struct lineFile *	lf
	)			[static]

Definition at line 19 of file dtdParse.c.

References errAbort(), lineFile::fileName, lineFile::lineIx, and nextWord().

Referenced by dtdParse(), parseAttribute(), parseElement(), and parseEntity().

{
char *word = nextWord(pLine);
if (word == NULL)
    errAbort("Missing data line %d of %s", lf->lineIx, lf->fileName);
return word;
}

Here is the call graph for this function:

Here is the caller graph for this function:

void needQuotedString	(	char *	in,
		char *	out,
		struct lineFile *	lf,
		char **	retNext
	)

Definition at line 28 of file dtdParse.c.

References errAbort(), lineFile::fileName, lineFile::lineIx, and parseQuotedString().

Referenced by parseAttribute(), and parseEntity().

{
if (!parseQuotedString(in, out, retNext))
    errAbort("Missing closing quote line %d of %s", lf->lineIx, lf->fileName);
}

Here is the call graph for this function:

Here is the caller graph for this function:

static void parseAttribute	(	char *	line,
		char *	textField,
		struct hash *	elHash,
		struct lineFile *	lf
	)			[static]

Definition at line 196 of file dtdParse.c.

References AllocVar, dtdElement::attributes, cloneString(), errAbort(), lineFile::fileName, hashFindVal(), lineFile::lineIx, mixedCaseName(), dtdAttribute::mixedCaseName, dtdAttribute::name, needNextWord(), needQuotedString(), dtdAttribute::required, sameString, sameWord, skipLeadingSpaces(), slAddTail(), TRUE, dtdAttribute::type, and dtdAttribute::usual.

Referenced by dtdParse().

{
char *word;
struct dtdAttribute *att;
struct dtdElement *el;
char *e;

/* Get rid of trailing '>' */
e = strrchr(line, '>');
if (e == NULL)
    errAbort("Missing '>' line %d of %s", lf->lineIx, lf->fileName);
*e = 0;

word = needNextWord(&line, lf);
if ((el = hashFindVal(elHash, word)) == NULL)
    errAbort("Undefined %s line %d of %s", word, lf->lineIx, lf->fileName);
word = needNextWord(&line, lf);
if (sameString(word, textField))
    errAbort("Name conflict with text field name line %d of %s", lf->lineIx, lf->fileName);
AllocVar(att);
att->name = cloneString(word);
att->mixedCaseName = mixedCaseName("", att->name);
word = needNextWord(&line, lf);
att->type = cloneString(word);
line = skipLeadingSpaces(line);
if (line[0] == '#')
    {
    word = needNextWord(&line, lf);
    if (sameWord("#REQUIRED", word))
        att->required = TRUE;
    else if (sameWord("#IMPLIED", word))
        att->usual = NULL;
    else
        errAbort("Unknown directive %s line %d of %s", word, lf->lineIx, lf->fileName);
    }
else if (line[0] == '\'' || line[0] == '"')
    {
    word = line;
    needQuotedString(word, word, lf, &line);
    att->usual = cloneString(word);
    }
else
    {
    word = needNextWord(&line, lf);
    att->usual = cloneString(word);
    }
slAddTail(&el->attributes, att);
}

Here is the call graph for this function:

Here is the caller graph for this function:

static struct dtdElement* parseElement	(	char *	prefix,
		char *	textField,
		char *	line,
		struct hash *	elHash,
		struct lineFile *	lf
	)			[static, read]

Definition at line 115 of file dtdParse.c.

References AllocVar, ArraySize, dtdElement::children, chopString(), cloneString(), dtdElChild::copyCode, errAbort(), lineFile::fileName, hashAddSaveName(), hashFindVal(), dtdElChild::isOr, lineFile::lineIx, dtdElement::lineIx, mixedCaseName(), dtdElement::mixedCaseName, dtdElChild::name, name, dtdElement::name, needNextWord(), sameString, slAddHead, slReverse(), and dtdElement::textType.

Referenced by dtdParse().

{
char *word, *s, *e;
char *words[256];
int wordCount, i;
struct dtdElChild *ec;
struct dtdElement *el;
boolean isOr;
char orCopyCode = '?';

word = needNextWord(&line, lf);
s = word + strlen(word)-1;
if (s[0] == '>')
   *s = 0;
if ((el = hashFindVal(elHash, word)) != NULL)
    errAbort("Duplicate element %s line %d and %d of %s", word, el->lineIx, lf->lineIx, lf->fileName);
AllocVar(el);
el->lineIx = lf->lineIx;
hashAddSaveName(elHash, word, el, &el->name);
el->mixedCaseName = mixedCaseName(prefix, el->name);
if (line != NULL && (s = strchr(line, '(')) != NULL)
    {
    s += 1;
    if ((e = strchr(line, ')')) == NULL)
        errAbort("Missing ')' line %d of %s", lf->lineIx, lf->fileName);
    *e = 0;
    isOr = (strchr(s, '|') != NULL);
    if (isOr)
      {
        orCopyCode = *(e+1);
        if ((orCopyCode != '+') && (orCopyCode != '*'))
          orCopyCode = '?';
      }
    wordCount = chopString(s, "| ,\t", words, ArraySize(words));
    if (wordCount == ArraySize(words))
        errAbort("Too many children in list line %d of %s", lf->lineIx, lf->fileName);
    for (i=0; i<wordCount; ++i)
        {
        char *name = words[i];
        int len = strlen(name);
        char lastC = name[len-1];
        if (name[0] == '#')
            {
            if (isOr)
                errAbort("# character in enumeration not allowed line %d of %s",
                   lf->lineIx, lf->fileName);
            if (el->textType != NULL)
                errAbort("Multiple types for text between tags line %d of %s", 
                        lf->lineIx, lf->fileName);
            el->textType = cloneString(name);
            }
        else
            {
            AllocVar(ec);
            slAddHead(&el->children, ec);
            ec->isOr = isOr;
            if (isOr)
               ec->copyCode = orCopyCode;
            else
                {
                if (lastC == '+' || lastC == '?' || lastC == '*')
                    {
                    ec->copyCode = lastC;
                    name[len-1] = 0;
                    }
                else
                    ec->copyCode = '1';
                }
            if (sameString(name, textField))
                errAbort("Name conflict with default text field name line %d of %s", lf->lineIx, lf->fileName);
            ec->name = cloneString(name);
            }
        }
    slReverse(&el->children);
    }
return el;
}

Here is the call graph for this function:

Here is the caller graph for this function:

void parseEntity	(	struct hash *	entityHash,
		struct hash *	predefEntityHash,
		char *	line,
		struct lineFile *	lf
	)

Definition at line 248 of file dtdParse.c.

References cloneString(), errAbort(), lineFile::fileName, hashAdd(), hashFindVal(), hashLookup(), lineFile::lineIx, name, needNextWord(), needQuotedString(), sameString, and skipLeadingSpaces().

Referenced by dtdParse().

{
char *percent = needNextWord(&line, lf);
char *name = needNextWord(&line, lf);
char *value = skipLeadingSpaces(line);
if (value[0] != '"')
    errAbort("Expecting quoted string at end of ENTITY tag line %d of %s",
        lf->lineIx, lf->fileName);
needQuotedString(value, value, lf, &line);
if (!sameString(percent, "%"))
    errAbort("Expecting %% after ENTITY tag line %d of %s", lf->lineIx, lf->fileName);
if (hashLookup(predefEntityHash, name) == NULL)
/* We don't want to overwrite the predefined entities.  These are all
 * defined to be #PCDATA or CDATA for the benefit of non-UCSC XML tools.
 * Internally we map them to #INT/#FLOAT etc. so we can have numbers
 * as well as strings in our C structures and relational database tables. */
    {
    char *oldVal = hashFindVal(entityHash, name);
    if (oldVal != NULL)
        {
        if (!sameString(oldVal, value))
            errAbort("Entity %s redefined line %d of %s", name, lf->lineIx, lf->fileName);
        }
    else
        {
        hashAdd(entityHash, name, cloneString(value));
        }
    }
}

Here is the call graph for this function:

Here is the caller graph for this function:

static void syntaxError

(

struct lineFile *

lf

)

[static]

Definition at line 13 of file dtdParse.c.

References errAbort(), lineFile::fileName, and lineFile::lineIx.

Referenced by dtdParse().

{
errAbort("Syntax error line %d of %s", lf->lineIx, lf->fileName);
}

Here is the call graph for this function:

Here is the caller graph for this function:

---