def printLine(self, line):
sty = style.currentStyle()
fontName = sty.font.name
fontSize = sty.fontsize
lineWidth = fitz.getTextlength(line, fontName, fontSize)
budget = width - margin - self.indent - self.insertPoint.x
if lineWidth < budget:
self.printSegment(line)
else:
prefix = line
suffix = line
while True:
index = line.rfind(" ", 0, len(prefix))
if index == -1: # No spaces found. Print it on the next line
self.cr(chr(0xAC))
self.printSegment(suffix)
break
prefix = line[:index]
prefixWidth = fitz.getTextlength(prefix,
fontname=fontName,
fontsize=fontSize)
if prefixWidth > budget: # Still too large. Chop again.
continue
else:
self.printSegment(prefix)
suffix = line[index + 1:]
# Recurse for the remainder. It might be too long as well.
self.cr(chr(0xAC))
self.printLine(suffix)
break
def createTextbox(textfile, textboxRect, page, path):
fontSize = 12
linecounter = 0
# Extracts text from file and eliminates blank lines
# Counts the number of lines in the file
with open(textfile) as input_txt:
lineString = ""
lineArray = []
for line in input_txt:
if not line.isspace():
lineString = lineString + line
lineArray.append(line)
linecounter = linecounter + 1
lineLength = fitz.getTextlength(line, "Times-Roman", fontSize)
while (lineLength > textboxRect.width):
linecounter = linecounter + 1
lineLength = lineLength - textboxRect.width
# At font size 12, textbox can hold up to 10 lines of text
# Decreases font size by increments of 4 while line capacity is exceeded
lineCapacity = 10
while (linecounter > lineCapacity):
# Line capacity of textbox increases by the percent increase from each font size decrement
# Ex: 12 --> 8 makes for a 50% increase in line capacity
percentIncrease = calculatePercentIncrease(fontSize, fontSize - 2)
lineCapacity = lineCapacity + (lineCapacity * percentIncrease)
# Decrement font size by increment of 2
fontSize = fontSize - 2
linecounter = 0
# Recalculate the number of lines
# This is necessary since decreasing font size may decrease the number of lines
# When text exceeds horizontal boundaries, a new line is started
for line in lineArray:
linecounter = linecounter + 1
lineLength = fitz.getTextlength(line, "Times-Roman", fontSize)
while (lineLength > textboxRect.width):
linecounter = linecounter + 1
lineLength = lineLength - textboxRect.width
# Generate textbox with text
textboxShape = page.new_shape()
textboxShape.draw_rect(textboxRect)
textboxShape.finish(color=getColor("dodgerblue4"),
fill_opacity=1,
closePath=False)
textboxShape.commit()
page.insertTextbox(textboxRect,
lineString,
fontsize=fontSize,
fontname="Times-Roman",
align=0)
return
def insert_fields(self, fields, form_type, PdfProcess):
for _, field in enumerate(fields):
document_id = field["document_id"]
value = field["value"].upper()
location = field["location"]
index = field["index"] if "index" in field else 0
rect = PdfProcess.find_text(document_id, index=index)
if rect is None:
continue
max_width = 0
if form_type == "con29r":
max_width = 200
elif form_type == "llc1":
max_width = 200
elif form_type == "con29o":
max_width = 175
point = (rect.x1 + 2, rect.y0)
if location == "below":
point = (rect.x0, rect.y0 + rect.height)
elif location == "before":
text_length = fitz.getTextlength(value, fontsize=9)
point = (rect.x1 - rect.width - text_length, rect.y0 - 1)
rect = fitz.Rect(point[0], point[1], point[0] + max_width,
point[1] + 200)
PdfProcess.insert_textbox(value, rect, fontsize=9)
def append_TOC(existingDoc, newDoc, entryname, filename, startPoint, path):
# Appends the new page to the existing document that contains the table of contents
existingDoc.insert_pdf(newDoc)
TOC_page = existingDoc.load_page(page_id=0)
# Generates the text entry for the new page
TOC_page.insertText(startPoint, entryname, fontname="helv", fontsize=16, rotate=0)
x_distance = (fitz.getTextlength(entryname, fontname="helv", fontsize=16)) + 105
targetPageNumber = existingDoc.page_count
entrynumber = " %i" % targetPageNumber
while(x_distance < 475):
dotLocation = fitz.Point(x_distance, startPoint.y)
TOC_page.insertText(dotLocation, ".", fontname="helv", fontsize=16, rotate=0)
x_distance = x_distance + 5
TOC_page.insertText(dotLocation, entrynumber, fontname="helv", fontsize=16, rotate=0)
# Creates the hyperlink for the newly appended page
# When the entry is clicked on in the Table of Contents, user is sent to that particular page
linkRect = Rect(100, startPoint.y-20, x_distance + 25, startPoint.y + 15)
newLink = TOC_page.insert_link({'kind': 1, 'from': linkRect, 'type': 'goto', 'page': targetPageNumber-1, 'to': fitz.Point(0, 0), 'zoom': 0.0})
# Inserts the page number on the bottom of the newly appended page
insertedPage = existingDoc.load_page(page_id=-1)
pageNumberPoint = fitz.Point(294, 830)
insertPageNumber = "%i" % targetPageNumber
insertedPage.insertText(pageNumberPoint, insertPageNumber, fontname="Times-Roman", fontsize=14, rotate=0)
# Calculates the new start point for the next entry and saves the pdf
newStartPoint = fitz.Point(100, startPoint.y + 35)
existingDoc.save(path + filename)
return newStartPoint
def printSegment(self, line):
sty = style.currentStyle()
fontName = sty.font.name
fontSize = sty.fontsize
lineWidth = fitz.getTextlength(line,
fontname=fontName,
fontsize=fontSize)
# self.currentPage.insertText(
# self.insertPoint,
# str(self.insertPoint.x),
# fontname="cobo",
# fontsize=fontSize * 2,
# )
log.debug(f"printSegment: {line}")
self.currentPage.insertText(self.insertPoint,
line,
fontname=fontName,
fontsize=fontSize)
if self.linkDestination:
self.linkRects.append(
fitz.Rect(
self.insertPoint.x,
self.insertPoint.y - lineheight,
self.insertPoint.x + lineWidth,
self.insertPoint.y,
))
self.insertPoint.x += lineWidth
def printRight(self, line):
sty = style.currentStyle()
lineWidth = fitz.getTextlength(line,
fontname=sty.font.name,
fontsize=sty.fontsize)
self.insertPoint.x = width - margin - lineWidth
self.currentPage.insertText(self.insertPoint,
line,
fontname=sty.font.name,
fontsize=sty.fontsize)
def softbreak(self, node=None, entering=None):
sty = style.currentStyle()
fontName = sty.font.name
fontSize = sty.fontsize
lineWidth = fitz.getTextlength(" ",
fontname=fontName,
fontsize=fontSize)
budget = width - margin - self.insertPoint.x
if lineWidth < budget:
self.printSegment(" ")
else:
pass # We're about to to a hard break anyway.
def apply(original, tokens: List[PDFToken], outfile, highlight=False):
pdf_original = fitz.open(str(original))
pdf_corrected = fitz.open()
PDFTokenizer.log.info('Copying images from original to corrected PDF')
for page in pdf_original:
PDFTokenizer.log.debug(f'(page {page.number})')
newpage = pdf_corrected.newPage(-1, page.rect.width,
page.rect.height)
for image_info in page.getImageList():
xref = image_info[0]
stream = pdf_original.extractImage(xref)['image']
newpage.insertImage(page.rect, stream=stream)
blue = fitz.utils.getColor('blue')
red = fitz.utils.getColor('red')
PDFTokenizer.log.info('Inserting tokens in corrected PDF')
for token in sorted(tokens, key=lambda x: x.ordering):
if token.is_discarded:
continue
page = pdf_corrected[token.ordering[0]]
word = token.gold or token.original
# Adjust rectangle to fit word:
fontfactor = 0.70
size = token.rect.height * fontfactor
textwidth = fitz.getTextlength(word, fontsize=size)
rect = fitz.Rect(
token.rect.x0, token.rect.y0,
max(token.rect.x1, token.rect.x0 + textwidth + 1.0),
token.rect.y1 + token.rect.height)
res = page.insertTextbox(rect,
f'{word} ',
fontsize=size,
render_mode=3)
if res < 0:
PDFTokenizer.log.warning(
f'Token was not inserted properly: {word}\n'
f' -- token.rect: {token.rect}\n'
f' -- rect: {rect}\n'
f' -- font size: {size}\n'
f' -- calc.width: {textwidth} rect.width: {rect.width}\n'
f' -- rect.height: {rect.height} result: {res}\n')
if highlight:
page.drawRect(rect, color=red)
elif highlight:
page.drawRect(rect, color=blue)
PDFTokenizer.log.info(f'Saving corrected PDF to {outfile}')
pdf_corrected.save(str(outfile)) #, garbage=4, deflate=True)
def generateTOC(existingDoc, filename, path):
generatedPage = existingDoc.newPage(pno=0)
# Generates the Table of Contents Title
tableOfContentsText = "-- Table of Contents --"
TOC_textLength = fitz.getTextlength(tableOfContentsText)
TOC_startPoint_X = ((595 / 2) - TOC_textLength)
TOC_startPoint_Y = 85
TOC_startPoint = fitz.Point(TOC_startPoint_X, TOC_startPoint_Y)
generatedPage.insertText(TOC_startPoint, tableOfContentsText, fontname="Times-Roman", color=(0, 0.35, 0.8),
fontsize=24, rotate=0)
# Inserts the page number at the bottom of the page.
# Table of Contents will be page 1
pageNumberPoint = fitz.Point(294, 815)
generatedPage.insertText(pageNumberPoint, "1", fontname="Times-Roman", fontsize=14, rotate=0)
existingDoc.save(path + filename)
return existingDoc
def apply(original,
tokens: List[HOCRToken],
outfile: Path,
highlight=False):
pdf = fitz.open()
pix = fitz.Pixmap(str(original))
page = pdf.newPage(-1, width=pix.width, height=pix.height)
page.insertImage(page.rect, pixmap=pix)
for token in progressbar.progressbar(tokens):
if token.is_discarded:
continue
page = pdf[token.page]
word = token.gold or token.original
# Adjust rectangle to fit word:
fontfactor = 0.70
size = token.rect.height * fontfactor
textwidth = fitz.getTextlength(word, fontsize=size)
rect = fitz.Rect(
token.rect.x0, token.rect.y0,
max(token.rect.x1, token.rect.x0 + textwidth + 1.0),
token.rect.y1 + token.rect.height * 2)
res = page.insertTextbox(rect,
f'{word} ',
fontsize=size,
color=(1, 0, 0))
if res < 0:
HOCRTokenizer.log.warning(
f'Token was not inserted properly: {word}\n'
f' -- token.rect: {token.rect}\n'
f' -- rect: {rect}\n'
f' -- font size: {size}\n'
f' -- calc.width: {textwidth} rect.width: {rect.width}\n'
f' -- rect.height: {rect.height} result: {res}\n')
pdf.save(str(outfile.with_suffix('.pdf')))
def insert_extra_info(extra, exam, test_mode=False, test_folder=None):
"""Creates the extra info (ususally student name and id) boxes and places them in the first page.
Arguments:
extra {dict} -- (Str:Str) dictioary with student id and name.
exam {fitz.Document} -- PDF document type returned as the exam, similar to a dictionary with the ge numbers as the keys.
Keyword Arguments:
test_mode {bool} -- Boolean elements used for testing, testing case with show the documents. (default: {False})
test_folder {Str} -- String for where to place the generated test files. (default: {None})
Raises:
ValueError: Raise error if the student name and number is not encodable.
Returns:
fitz.Document -- The same exam object as the input, except we add the extra infor into the first page.
"""
# Get page width and height
page_width = exam[0].bound().width
page_height = exam[0].bound().height
student_id = extra["id"]
student_name = extra["name"]
# a file for the student-details
YSHIFT = 0.4 # where on page is centre of box 0=top, 1=bottom
# Creating the student id \n name text file
txt = "{}\n{}".format(student_id, student_name)
sign_here = "Please sign here"
# Getting the dimentions of the box
try:
student_id_width = (max(
fitz.get_text_length(student_id, fontsize=36,
fontname="Helvetica"),
fitz.get_text_length(
student_name, fontsize=36, fontname="Helvetica"),
fitz.get_text_length(sign_here, fontsize=48, fontname="Helvetica"),
) * 1.1 * 0.5)
except AttributeError:
# workaround https://github.com/pymupdf/PyMuPDF/issues/1085
# TODO: drop this code when we dependent on PyMuPDF>=1.18.14
student_id_width = (max(
fitz.getTextlength(student_id, fontsize=36, fontname="Helvetica"),
fitz.getTextlength(student_name, fontsize=36,
fontname="Helvetica"),
fitz.getTextlength(sign_here, fontsize=48, fontname="Helvetica"),
) * 1.1 * 0.5)
student_id_height = 36 * 1.3
# We have 2 rectangles for the student name and student id
student_id_rect_1 = fitz.Rect(
page_width // 2 - student_id_width,
page_height * YSHIFT - student_id_height,
page_width // 2 + student_id_width,
page_height * YSHIFT + student_id_height,
)
student_id_rect_2 = fitz.Rect(
student_id_rect_1.x0,
student_id_rect_1.y1,
student_id_rect_1.x1,
student_id_rect_1.y1 + 48 * 1.3,
)
exam[0].draw_rect(student_id_rect_1,
color=[0, 0, 0],
fill=[1, 1, 1],
width=2)
exam[0].draw_rect(student_id_rect_2,
color=[0, 0, 0],
fill=[1, 1, 1],
width=2)
# TODO: This could be put into one function
# Also VALA doesn't understand the TODO s
if is_possible_to_encode_as(txt, "Latin-1"):
fontname = "Helvetica"
elif is_possible_to_encode_as(txt, "gb2312"):
# TODO: Double-check encoding name. Add other CJK (how does Big5
# vs GB work?). Check printers can handle these or do we need to
# embed a font? (Adobe Acrobat users need to download something)
fontname = "china-ss"
else:
# TODO: or warn use Helvetica, get "?" chars
raise ValueError(
"Don't know how to write name {} into PDF".format(txt))
# We insert the student id text boxes
insertion_confirmed = exam[0].insert_textbox(
student_id_rect_1,
txt,
fontsize=36,
color=[0, 0, 0],
fontname=fontname,
fontfile=None,
align=1,
)
# TODO: VALA suggests we do the insertion_confirmed check here as well
assert (insertion_confirmed >
0), "Text didn't fit: shortname too long? or font issue/bug?"
# We insert the student name text boxes
insertion_confirmed = exam[0].insert_textbox(
student_id_rect_2,
sign_here,
fontsize=48,
color=[0.9, 0.9, 0.9],
fontname="Helvetica",
fontfile=None,
align=1,
)
# TODO: VALA suggests we do the insertion_confirmed check here as well
assert (insertion_confirmed >
0), "Text didn't fit: shortname too long? or font issue/bug?"
return exam
print("PDF conversion results for file '%s':" % (ifn,))
print(out_ctr, "lines read,", total_ctr,
"lines written,", nlines, "lines per page.")
print(ofn, "contains", len(doc), "pages.")
# Now add some header and footer to each created page
hdr_fontsz = 16 # header fontsize
ftr_fontsz = 8 # footer fontsize
blue = (0, 0, 1) # header / footer color
pspace = 500 # available line width
for page in doc:
footer = "%i (%i)" % (page.number + 1, len(doc)) # footer text
plen_ftr = fitz.getTextlength(footer, fontname="Helvetica", fontsize=ftr_fontsz)
page.insertText(fitz.Point(50, 50), ifn, # header = input filename
color = blue,
fontsize = hdr_fontsz)
page.drawLine(fitz.Point(50,60), fitz.Point(50 + pspace,60), # line below hdr
color= blue, width = 0.5)
page.drawLine(fitz.Point(50, height-33), # line above footer
fitz.Point(50 + pspace, height - 33),
color = blue, width = 0.5)
page.insertText(fitz.Point(50 + pspace - plen_ftr, # insert footer
height - 33 + ftr_fontsz * 1.2),
footer, fontsize = ftr_fontsz, color = blue)
def generateDoc(title, data, dataStart, dataEnd, bio1, bio2, bio3, bio4,
vid_filename, textfile, filename, pageNum, path):
# Creates a new blank PDF
doc = fitz.open()
generatedPage = doc.newPage()
font = "Times-Roman"
fontSize = 24
titleLength = fitz.getTextlength(title, font, fontSize)
# Prints the dimensions of the newly generated page.
# These values may be useful for determining the locations of the plots
pageRect = generatedPage.bound()
page_x0 = pageRect.x0
page_x1 = pageRect.x1
# Ensures that the title will always be centered, despite text length
pageMidpoint_X = (page_x1 - page_x0) / 2
titleStartPoint_X = pageMidpoint_X - (titleLength / 2)
titleStartPoint_Y = fontSize + 11
titleStartPoint = fitz.Point(titleStartPoint_X, titleStartPoint_Y)
generatedPage.insertText(titleStartPoint,
title,
fontname=font,
fontsize=fontSize,
rotate=0)
# Autogenerates the biometric plots
bio1_filename = path + "page%i_" % pageNum + bio1 + ".png"
bioPlotter.plotBiometric(data, dataStart, dataEnd, bio1, bio1_filename)
bio2_filename = path + "page%i_" % pageNum + bio2 + ".png"
bioPlotter.plotBiometric(data, dataStart, dataEnd, bio2, bio2_filename)
bio3_filename = path + "page%i_" % pageNum + bio3 + ".png"
bioPlotter.plotBiometric(data, dataStart, dataEnd, bio3, bio3_filename)
bio4_filename = path + "page%i_" % pageNum + bio4 + ".png"
bioPlotter.plotBiometric(data, dataStart, dataEnd, bio4, bio4_filename)
# Autogenerates the EEG heatmaps
eeg.eeg_viz(data, dataStart, dataEnd, path + "page%i_eeg_" % pageNum)
fontSize = 14
# Extracts a frame from the video in the specified time range
# extracted_frame_filename = path + "page%i_extracted_frame.jpg"% pageNum
# vidFrame.extractFrame(vid_filename, dataStart, dataEnd, extracted_frame_filename)
# Inserts heatmap visualizations
heatmapAlpha_Location = fitz.Rect(10, 50, 198, 238)
generatedPage.insertImage(heatmapAlpha_Location,
filename=path + "page%i_eeg_alpha.png" % pageNum,
keep_proportion=False)
alphaText = "Alpha Band"
textLength = fitz.getTextlength(alphaText, font, fontSize)
startPoint = fitz.Point(((10 + 94) - textLength / 2), 240)
generatedPage.insertText(startPoint,
alphaText,
fontname=font,
fontsize=fontSize,
rotate=0)
heatmapBeta_Location = fitz.Rect(203, 50, 391, 238)
generatedPage.insertImage(heatmapBeta_Location,
filename=path + "page%i_eeg_beta.png" % pageNum,
keep_proportion=False)
betaText = "Beta Band"
textLength = fitz.getTextlength(alphaText, font, fontSize)
startPoint = fitz.Point(((203 + 94) - textLength / 2), 240)
generatedPage.insertText(startPoint,
betaText,
fontname=font,
fontsize=fontSize,
rotate=0)
heatmapTheta_Location = fitz.Rect(396, 50, 585, 238)
generatedPage.insertImage(heatmapTheta_Location,
filename=path + "page%i_eeg_theta.png" % pageNum,
keep_proportion=False)
thetaText = "Theta Band"
textLength = fitz.getTextlength(alphaText, font, fontSize)
startPoint = fitz.Point(((396 + 94) - textLength / 2), 240)
generatedPage.insertText(startPoint,
thetaText,
fontname=font,
fontsize=fontSize,
rotate=0)
# Inserts biometric plots
bio3_Location = fitz.Rect(10, 443, 300, 653)
generatedPage.insertImage(bio3_Location,
filename=bio3_filename,
keep_proportion=False)
bio4_Location = fitz.Rect(305, 443, 595, 653)
generatedPage.insertImage(bio4_Location,
filename=bio4_filename,
keep_proportion=False)
bio1_Location = fitz.Rect(10, 253, 300, 463)
generatedPage.insertImage(bio1_Location,
filename=bio1_filename,
keep_proportion=False)
bio2_Location = fitz.Rect(305, 253, 595, 463)
generatedPage.insertImage(bio2_Location,
filename=bio2_filename,
keep_proportion=False)
# Generates textbox
textboxBack_Location = fitz.Rect(250, 650, 585, 815)
textHandler.createTextbox(textfile, textboxBack_Location, generatedPage,
path)
# Inserts video frame
vidFrame_Location = fitz.Rect(25, 675, 245, 799)
# generatedPage.insertImage(vidFrame_Location, filename=extracted_frame_filename, keep_proportion=False)
generatedPage.insertImage(vidFrame_Location,
filename=vid_filename,
keep_proportion=False)
# Saves the PDF -- not needed anymore
#doc.save(filename)
return doc
for block in image_blocks:
outpage.insertImage(block["bbox"], stream=block["image"])
print("Inserted an image on page", page.number)
for block in text_blocks: # read text blocks
shape.drawRect(block["bbox"]) # draw all text on white background,
# because images may cover same area
for line in block["lines"]: # for each line in the block ...
for span in line["spans"]: # for each span in the line ...
fontname = get_font(span["font"]) # get replacing fontname
fontsize = span["size"]
text = span["text"]
bbox = fitz.Rect(span["bbox"]) # text rectangle on input
text_size = fitz.getTextlength( # measure text length on output
text, fontname=fontname, fontsize=fontsize
)
# adjust fontsize if text is too long with new the font
if text_size > bbox.width:
fontsize *= bbox.width / text_size
try:
shape.insertText( # copy text to output page
bbox.bl, # insertion point on output page
text, # the text to insert
fontsize=fontsize, # fontsize
# decide on output font here: the place for sophistication!
fontname=fontname,
color=pdf_color(span["color"]),
)
except ValueError: