public EdifactModel Read(StreamReader sr,
EdifactParseSettings defaultSettings)
{
var model = new EdifactModel(); /* the current model, start with a new edifact model */
var segment = new EdifactSegmentModel(); /* the current segment, start with a new segment */
var dataElement = new EdifactDataElement(); /* the current dataElement, start with a new data element */
bool isEscaped = false; /* true, if the last character was a ReleaseCharacter */
ParserState state = ParserState.Tag; /* start in SegmentTag mode */
StringBuilder textbuffer = new StringBuilder(); /* text buffer */
/* try to read the parse settings by looking for the UNA element at the start of the stream */
Queue <char> unprocessedChars;
EdifactParseSettings settings = ReadUNA(sr, out unprocessedChars);
model.UNA = settings; /* store the UNA parse settings of the model */
if (settings == null) /* if the stream does not contain settings (it is missing the UNA segment) */
{
settings = defaultSettings; /* then use the provided defaultSettings for parsing the model */
}
int r;
/* start with the unprocessedChars from the ReadUNA read attempt, then switch to the rest of the stream */
while ((r = (unprocessedChars.Count > 0 ? unprocessedChars.Dequeue() : sr.Read())) != -1)
{
char c = (char)r;
bool add = false;
if (isEscaped) /* if the previous character was an escape */
{
add = true; /* then add it to the text buffer */
isEscaped = false; /* and stop the escaping */
}
else
{
if (c == settings.ReleaseCharacter) /* if it's an escape character, escape the next char */
{
isEscaped = true;
}
else if (c == settings.SegmentTagDataElementSeparator) /* if segment tag / data element change occurs */
{
switch (state)
{
case ParserState.Tag:
segment.SegmentType = textbuffer.Cut(); /* set the collected text as the segments type and reset the buffer */
state = ParserState.DataElement; /* the next separators will separate data elements */
break;
case ParserState.DataElement:
dataElement.Components.Add(textbuffer.Cut()); /* complete the component */
segment.DataElements.Add(dataElement); /* add the completed dataElement to segment */
dataElement = new EdifactDataElement(); /* start a new dataElement */
break;
}
}
else if (c == settings.ComponentDataElementSeparator)
{
dataElement.Components.Add(textbuffer.Cut()); /* complete the component */
}
else if (c == settings.SegmentTerminator) /* if it's the end of the segment */
{
if (segment.SegmentType == null) /* complete the segment if it's not completed yet */
{
segment.SegmentType = textbuffer.Cut();
}
else
{
dataElement.Components.Add(textbuffer.Cut()); /* complete the component */
segment.DataElements.Add(dataElement); /* add the completed dataElement to the segment */
dataElement = new EdifactDataElement(); /* and start a new dataElement */
}
model.Segments.Add(segment); /* add the completed segment to the model */
segment = new EdifactSegmentModel(); /* and start a new segment */
state = ParserState.Tag; /* switch back to the SegmentTag state */
}
else if (c == '\r' || c == '\n')
{
} /* ignore line breaks */
else
{
add = true; /* if this was no special character, add it to the buffer */
}
}
if (add)
{
textbuffer.Append(c); /* collect it in the text buffer */
}
}
return(model);
}
public EdifactModel Read(StreamReader sr,
EdifactParseSettings defaultSettings)
{
var model = new EdifactModel(); /* the current model, start with a new edifact model */
var segment = new EdifactSegmentModel(); /* the current segment, start with a new segment */
var dataElement = new EdifactDataElement(); /* the current dataElement, start with a new data element */
bool isEscaped = false; /* true, if the last character was a ReleaseCharacter */
ParserState state = ParserState.Tag; /* start in SegmentTag mode */
StringBuilder textbuffer = new StringBuilder(); /* text buffer */
/* try to read the parse settings by looking for the UNA element at the start of the stream */
Queue<char> unprocessedChars;
EdifactParseSettings settings = ReadUNA(sr, out unprocessedChars);
model.UNA = settings; /* store the UNA parse settings of the model */
if (settings == null) /* if the stream does not contain settings (it is missing the UNA segment) */
settings = defaultSettings; /* then use the provided defaultSettings for parsing the model */
int r;
/* start with the unprocessedChars from the ReadUNA read attempt, then switch to the rest of the stream */
while ( (r = (unprocessedChars.Count > 0 ? unprocessedChars.Dequeue() : sr.Read())) != -1)
{
char c = (char)r;
bool add = false;
if (isEscaped) /* if the previous character was an escape */
{
add = true; /* then add it to the text buffer */
isEscaped = false; /* and stop the escaping */
}
else
{
if (c == settings.ReleaseCharacter) /* if it's an escape character, escape the next char */
isEscaped = true;
else if (c == settings.SegmentTagDataElementSeparator) /* if segment tag / data element change occurs */
{
switch (state)
{
case ParserState.Tag:
segment.SegmentType = textbuffer.Cut(); /* set the collected text as the segments type and reset the buffer */
state = ParserState.DataElement; /* the next separators will separate data elements */
break;
case ParserState.DataElement:
dataElement.Components.Add(textbuffer.Cut()); /* complete the component */
segment.DataElements.Add(dataElement); /* add the completed dataElement to segment */
dataElement = new EdifactDataElement(); /* start a new dataElement */
break;
}
}
else if (c == settings.ComponentDataElementSeparator)
{
dataElement.Components.Add(textbuffer.Cut()); /* complete the component */
}
else if (c == settings.SegmentTerminator) /* if it's the end of the segment */
{
if (segment.SegmentType == null) /* complete the segment if it's not completed yet */
segment.SegmentType = textbuffer.Cut();
else
{
dataElement.Components.Add(textbuffer.Cut()); /* complete the component */
segment.DataElements.Add(dataElement); /* add the completed dataElement to the segment */
dataElement = new EdifactDataElement(); /* and start a new dataElement */
}
model.Segments.Add(segment); /* add the completed segment to the model */
segment = new EdifactSegmentModel(); /* and start a new segment */
state = ParserState.Tag; /* switch back to the SegmentTag state */
}
else if (c == '\r' || c == '\n') { } /* ignore line breaks */
else add = true; /* if this was no special character, add it to the buffer */
}
if (add)
textbuffer.Append(c); /* collect it in the text buffer */
}
return model;
}