~/

python -c "import pdfplumber; p='/tmp/workspace/attachments/a4a929c2-1f21-4680-bb3c-8338da8ead16/DOC-20261016-WA0005.pdf'; pdf=pdfplumber.open(p); print('PAGES',len(pdf.pages)); [print('--- PAGE',i+1,'---\n'+(page.extract_text() or '')) for i,page in enumerate(pdf.pages)]"

mkdir -p "/tmp/workspace/esl-handbook/assets" "/tmp/workspace/esl-handbook/output" "/tmp/workspace/esl-handbook/src" && python -c "import fitz; p='/tmp/workspace/attachments/a4a929c2-1f21-4680-bb3c-8338da8ead16/DOC-20261016-WA0005.pdf'; d=fitz.open(p); [pg.get_pixmap(matrix=fitz.Matrix(2,2), alpha=False).save(f'/tmp/workspace/esl-handbook/assets/source_page_{i+1}.png') for i,pg in enumerate(d)]" && ls -lh "/tmp/workspace/esl-handbook/assets"

~/esl-handbook

python -c "import importlib.util; print('pymupdf',importlib.util.find_spec('pymupdf')); print('pdf2image',importlib.util.find_spec('pdf2image')); print('pypdfium2',importlib.util.find_spec('pypdfium2')); print('PIL',importlib.util.find_spec('PIL')); print('easyocr',importlib.util.find_spec('easyocr')); print('pytesseract',importlib.util.find_spec('pytesseract'))"

python -c "import pypdfium2 as pdfium; p='/tmp/workspace/attachments/a4a929c2-1f21-4680-bb3c-8338da8ead16/DOC-20261016-WA0005.pdf'; doc=pdfium.PdfDocument(p); [page.render(scale=2.5).to_pil().save(f'/tmp/workspace/esl-handbook/assets/source_page_{i+1}.png') for i,page in enumerate(doc)]" && ls -lh "/tmp/workspace/esl-handbook/assets"

python -c "from PIL import Image; import glob,os; [Image.open(p).convert('RGB').resize((850,int(Image.open(p).height*850/Image.open(p).width))).save(p.replace('.png','_small.jpg'),quality=72,optimize=True) for p in glob.glob('/tmp/workspace/esl-handbook/assets/source_page_*.png')]" && ls -lh "/tmp/workspace/esl-handbook/assets"

which tesseract || true; python -c "from PIL import Image; im=Image.open('/tmp/workspace/esl-handbook/assets/source_page_1_small.jpg'); print(im.size)"

python -c "import importlib.util; mods=['keras_ocr','rapidocr_onnxruntime','paddleocr','doctr','cv2']; [print(m,importlib.util.find_spec(m)) for m in mods]"

Inspect the attached scanned PDF at /tmp/workspace/attachments/a4a929c2-1f21-4680-bb3c-8338da8ead16/DOC-20261016-WA0005.pdf. Extract all readable English source text and vocabulary exactly enough to write an ESL handbook. It has 3 pages and is image-only. You may install/use OCR libraries if needed. Return a concise structured list of all words, sentences, and themes found. Do not create a file.

~/attachments/a4a929c2-1f21-4680-bb3c-8338da8ead16

python -c "import fitz; p='/tmp/workspace/attachments/a4a929c2-1f21-4680-bb3c-8338da8ead16/DOC-20261016-WA0005.pdf'; d=fitz.open(p); print('pages',len(d)); [page.get_pixmap(matrix=fitz.Matrix(3,3), alpha=False).save('/tmp/page_%d.png'%(i+1)) for i,page in enumerate(d)]" && python -c "import pytesseract; print(pytesseract.get_tesseract_version())"

python -m pip install --quiet pymupdf easyocr && python -c "import fitz; print(fitz.__doc__.split()[1])"

python -c "import pymupdf; p='/tmp/workspace/attachments/a4a929c2-1f21-4680-bb3c-8338da8ead16/DOC-20261016-WA0005.pdf'; d=pymupdf.open(p); print(len(d)); [pg.get_pixmap(matrix=pymupdf.Matrix(3,3), alpha=False).save('/tmp/page_%d.png'%(i+1)) for i,pg in enumerate(d)]" && ls -lh /tmp/page_*.png

python - <<'PY' import easyocr r=easyocr.Reader(['en'],gpu=False,verbose=False) for i in range(1,4): print('\n===== PAGE',i,'=====') out=r.readtext('/tmp/page_%d.png'%i,detail=1,paragraph=False) for b,t,c in out: print(f'{c:.2f}\t{t}') PY

python - <<'PY' from PIL import Image for i in range(1,4): im=Image.open(f'/tmp/page_{i}.png') print(i,im.size) im.thumbnail((1400,1800)) im.convert('RGB').save(f'/tmp/small_{i}.jpg',quality=70,optimize=True) PY ls -lh /tmp/small_*.jpg

which tesseract || true; python - <<'PY' from PIL import Image import easyocr r=easyocr.Reader(['en'],gpu=False,verbose=False) for i in range(1,4): im=Image.open('/tmp/page_%d.png'%i).convert('RGB') im.thumbnail((1200,1600)) im.save('/tmp/ocr_%d.jpg'%i,quality=95) print('\n===== PAGE',i,'=====') for _,t,c in r.readtext('/tmp/ocr_%d.jpg'%i,detail=1,paragraph=False,canvas_size=1280): print(f'{c:.2f}\t{t}') PY

python - <<'PY' from PIL import Image import easyocr r=easyocr.Reader(['en'],gpu=False,verbose=False) for pg in range(1,4): im=Image.open(f'/tmp/page_{pg}.png').convert('RGB') W,H=im.size for y in range(0,H,1100): cr=im.crop((0,y,W,min(y+1250,H))) cr.thumbnail((1400,600)) print(f'\n===== P{pg}, Y{y} =====') for _,t,c in r.readtext(cr,detail=1,paragraph=False,canvas_size=1600): print(f'{c:.2f}\t{t}') PY

python - <<'PY' from PIL import Image import numpy as np, easyocr r=easyocr.Reader(['en'],gpu=False,verbose=False) for pg in range(1,4): im=Image.open(f'/tmp/page_{pg}.png').convert('RGB'); W,H=im.size for y in range(0,H,1200): cr=im.crop((0,y,W,min(y+1350,H))); cr.thumbnail((1500,700)) print(f'\n===== P{pg}, Y{y} =====') for _,t,c in r.readtext(np.array(cr),detail=1,paragraph=False,canvas_size=1600): print(f'{c:.2f}\t{t}') PY

fc-match 'DejaVu Sans' -f '%{file}\n' | head -1

cat > "/tmp/workspace/esl-handbook/src/create_handbook.py" <<'PY' from reportlab.lib.pagesizes import A4 from reportlab.lib import colors from reportlab.lib.colors import HexColor from reportlab.pdfbase import pdfmetrics from reportlab.pdfbase.ttfonts import TTFont from reportlab.pdfgen import canvas from reportlab.lib.utils import simpleSplit from reportlab.pdfbase.pdfmetrics import stringWidth import os OUT='/tmp/workspace/esl-handbook/output/Present_Tenses_Conversation_Handbook.pdf' REG='/usr/share/fonts/google-noto-vf/NotoSans[wght].ttf' BOLD='/usr/share/fonts/truetype/dejavu/DejaVuSans-Bold.ttf' pdfmetrics.registerFont(TTFont('Noto', REG)) pdfmetrics.registerFont(TTFont('NotoBold', BOLD)) W,H=A4 navy=HexColor('#16324F'); blue=HexColor('#377DFF'); sky=HexColor('#EAF4FF'); mint=HexColor('#E7F8F2'); yellow=HexColor('#FFF4B8'); coral=HexColor('#FF8370'); purple=HexColor('#7A5AF8'); dark=HexColor('#19324A'); gray=HexColor('#587083') items=[ ('What do you see?','Что ты видишь?','I see ice.','Я вижу лёд.'), ('What is the weather like?','Какая погода?','It is cold and rainy.','Холодно и дождливо.'), ('Is water a liquid?','Вода - жидкость?','Yes, it is a liquid.','Да, это жидкость.'), ('Is a book a solid?','Книга - твёрдое тело?','Yes, it is a solid.','Да, это твёрдое тело.'), ('Do you put ice in drinks?','Ты кладёшь лёд в напитки?','Yes, I do. Ice makes drinks cold.','Да. Лёд делает напитки холодными.'), ('Is the ice melting?','Лёд тает?','Yes, it is melting.','Да, он тает.'), ('Is water flowing in the river?','Вода течёт в реке?','Yes, it is flowing over stones.','Да, она течёт по камням.'), ('Do you see steam on hot water?','Ты видишь пар над горячей водой?','Yes, I do.','Да.'), ('Is steam a gas?','Пар - это газ?','Yes, it is a gas.','Да, это газ.'), ('Does water freeze on a very cold day?','Вода замерзает в очень холодный день?','Yes, it does.','Да.'), ('Is ice cream melting in the sun?','Мороженое тает на солнце?','Yes, it is melting.','Да, оно тает.'), ('Are you watering the plants?','Ты поливаешь растения?','Yes, I am.','Да.'), ('Are the plants growing?','Растения растут?','Yes, they are growing.','Да, они растут.'), ('Do children go to school when there is a lot of snow?','Дети ходят в школу, когда много снега?','No, they do not. They stay home.','Нет. Они остаются дома.'), ('Are people slipping on the ice?','Люди скользят на льду?','Sometimes they are.','Иногда они скользят.'), ] vocab=[ ('flow','to move like water in a river','течь, течь потоком'), ('solid','a thing with a fixed shape, like a book','твёрдое тело; твёрдый'), ('liquid','a form of water that can flow','жидкость; жидкий'), ('gas','a form of matter that we cannot hold in our hands','газ'), ('heat','to make something hot','нагревать'), ('steam','a gas from very hot water','пар'), ('ice','frozen water','лёд'), ('freeze','to change water into ice','замерзать; замораживать'), ('melt','to change from ice to liquid','таять; растапливать'), ] def box(c,x,y,w,h,fill, radius=14): c.setFillColor(fill); c.roundRect(x,y,w,h,radius,fill=1,stroke=0) def text(c,s,x,y,size=11,font='Noto',color=dark): c.setFont(font,size); c.setFillColor(color); c.drawString(x,y,s) def wrap(c,s,x,y,width,size=11,font='Noto',color=dark,leading=None): leading=leading or size*1.35; lines=simpleSplit(s,font,size,width); c.setFillColor(color); c.setFont(font,size) for ln in lines: c.drawString(x,y,ln); y-=leading return y def header(c,title,subtitle,page): c.setFillColor(sky); c.rect(0,H-91,W,91,fill=1,stroke=0) c.setFillColor(blue); c.circle(W-38,H-28,32,fill=1,stroke=0) c.setFillColor(coral); c.circle(W-73,H-76,17,fill=1,stroke=0) text(c,title,40,H-48,22,'NotoBold',navy); text(c,subtitle,40,H-71,10,'Noto',gray) text(c,f'{page}',W-48,25,9,'NotoBold',gray) def qcard(c,idx, item, top, hue): q,qr,a,ar=item; x=38; w=W-76; h=130; y=top-h box(c,x,y,w,h,HexColor('#FFFFFF')) c.setFillColor(hue); c.circle(x+21,y+h-22,14,fill=1,stroke=0) c.setFillColor(colors.white); c.setFont('NotoBold',10); c.drawCentredString(x+21,y+h-25.5,str(idx)) text(c,'QUESTION',x+45,y+h-22,8,'NotoBold',hue) yy=wrap(c,q,x+45,y+h-41,w-60,12.3,'NotoBold',navy,15) yy=wrap(c,qr,x+45,yy-2,w-60,10.3,'Noto',gray,13) c.setStrokeColor(HexColor('#DCE8F4')); c.line(x+45,y+48,x+w-18,y+48) text(c,'ANSWER',x+45,y+35,8,'NotoBold',purple) yy=wrap(c,a,x+45,y+19,w-60,11,'NotoBold',dark,14) wrap(c,ar,x+45,yy-1,w-60,10.1,'Noto',gray,12) return y c=canvas.Canvas(OUT,pagesize=A4) c.setTitle('Present Simple & Present Continuous - ESL Conversation Handbook') # Cover c.setFillColor(HexColor('#F5FAFF')); c.rect(0,0,W,H,fill=1,stroke=0) for x,y,r,col in [(65,740,31,yellow),(515,740,40,mint),(540,130,52,HexColor('#E8DEFF')),(70,155,39,HexColor('#FFE6DF'))]: c.setFillColor(col);c.circle(x,y,r,fill=1,stroke=0) box(c,44,175,W-88,475,colors.white,24) text(c,'SPEAK & LEARN',75,605,13,'NotoBold',blue) text(c,'Present Simple &',75,548,29,'NotoBold',navy) text(c,'Present Continuous',75,510,29,'NotoBold',navy) wrap(c,'15 conversation questions and answers about water, weather, and changes.',75,466,370,14,'Noto',gray,19) # decorative ice/water c.setFillColor(blue); c.roundRect(400,495,82,82,15,fill=1,stroke=0); c.setFillColor(HexColor('#BAE5FF')); c.circle(442,536,24,fill=1,stroke=0) c.setFillColor(mint); c.roundRect(75,292,390,112,15,fill=1,stroke=0) text(c,'Practice rule',98,368,12,'NotoBold',purple) wrap(c,'Read the English sentence aloud. Then read the Russian translation. Use the model answer or make your own answer.',98,340,335,11,'Noto',dark,15) text(c,'Grade 3 ESL • Source topic: Forms and States',75,220,10,'NotoBold',gray) text(c,'English • Русский',75,198,12,'NotoBold',blue) c.showPage() # Conversation pages for pg in range(4): header(c,'Conversation Practice', 'Present simple • Present continuous • Read, ask, answer', pg+2) start=H-116 for k in range(4): idx=pg*4+k if idx < len(items): start=qcard(c,idx+1,items[idx],start,[blue,coral,purple,HexColor('#27A77D')][k]) - 12 c.setFillColor(gray); c.setFont('Noto',8.5); c.drawCentredString(W/2,38,'Say the question. Listen to the answer. Then switch roles!') c.showPage() # vocab 1 header(c,'Vocabulary Builder', 'Words from the source material • English meaning, then Russian meaning',6) text(c,'Word bank',40,H-115,17,'NotoBold',navy) text(c,'Say each word. Read its meaning. Use it in a short answer.',40,H-137,10.5,'Noto',gray) def vcard(c,word,en,ru,top,color): x=39; y=top-96; w=W-78 box(c,x,y,w,96,colors.white) box(c,x+13,y+19,112,58,color,11) c.setFillColor(colors.white);c.setFont('NotoBold',15);c.drawCentredString(x+69,y+52,word) text(c,'English meaning',x+145,y+70,8,'NotoBold',blue); wrap(c,en,x+145,y+52,w-165,10.2,'Noto',dark,13) text(c,'Русское значение',x+145,y+27,8,'NotoBold',purple); wrap(c,ru,x+145,y+9,w-165,10.2,'Noto',gray,13) return y start=H-155 for i,v in enumerate(vocab[:5]): start=vcard(c,*v,start,[blue,coral,purple,HexColor('#31A47B'),HexColor('#E5A814')][i])-12 c.showPage() # vocab 2 header(c,'Vocabulary Builder', 'Words from the source material • English meaning, then Russian meaning',7) text(c,'Word bank',40,H-115,17,'NotoBold',navy) text(c,'Review: point, say, and explain each word.',40,H-137,10.5,'Noto',gray) start=H-155 for i,v in enumerate(vocab[5:]): start=vcard(c,*v,start,[blue,coral,purple,HexColor('#31A47B')][i])-16 box(c,40,70,W-80,60,yellow,15) text(c,'Remember!',60,105,11,'NotoBold',navy) wrap(c,'A solid, a liquid, and a gas are different forms of matter.',60,84,W-120,10.5,'Noto',dark,13) c.save() print(OUT) PY python "/tmp/workspace/esl-handbook/src/create_handbook.py" && ls -lh "/tmp/workspace/esl-handbook/output/Present_Tenses_Conversation_Handbook.pdf"

fc-match 'Noto Sans:style=Bold' -f '%{file}\n' | head -1 && python -c "p='/tmp/workspace/esl-handbook/src/create_handbook.py'; s=open(p).read(); s=s.replace(\"BOLD='/usr/share/fonts/truetype/dejavu/DejaVuSans-Bold.ttf'\", \"BOLD='/usr/share/fonts/google-noto-vf/NotoSans[wght].ttf'\"); open(p,'w').write(s)" && python "/tmp/workspace/esl-handbook/src/create_handbook.py" && ls -lh "/tmp/workspace/esl-handbook/output/Present_Tenses_Conversation_Handbook.pdf"

~/

python - <<'PY' p='/tmp/workspace/esl-handbook/src/create_handbook.py' s=open(p,encoding='utf-8').read() start=s.index('items=[') end=s.index(']\nvocab=',start)+1 new='''items=[ ('What do you see in the picture?','Что ты видишь на картинке?','I see water, ice, and steam.','Я вижу воду, лёд и пар.'), ('Is water a solid or a liquid?','Вода - твёрдое тело или жидкость?','Water is a liquid.','Вода - жидкость.'), ('What happens to water on a very cold day?','Что происходит с водой в очень холодный день?','Water freezes.','Вода замерзает.'), ('Is the water freezing now?','Вода сейчас замерзает?','Yes, it is freezing.','Да, она замерзает.'), ('What do you put in drinks to make them cold?','Что ты кладёшь в напитки, чтобы они стали холодными?','I put ice in drinks.','Я кладу лёд в напитки.'), ('Is ice a solid?','Лёд - это твёрдое тело?','Yes, ice is a solid.','Да, лёд - это твёрдое тело.'), ('What happens to ice cream on a hot day?','Что происходит с мороженым в жаркий день?','Ice cream melts.','Мороженое тает.'), ('Is the ice cream melting in the sun?','Мороженое тает на солнце?','Yes, it is melting in the sun.','Да, оно тает на солнце.'), ('What can you see on very hot water?','Что ты можешь увидеть над очень горячей водой?','I can see steam.','Я могу увидеть пар.'), ('Is steam a liquid or a gas?','Пар - это жидкость или газ?','Steam is a gas.','Пар - это газ.'), ('What does water do in a river?','Что делает вода в реке?','Water flows over stones.','Вода течёт по камням.'), ('Is the river water flowing now?','Вода в реке сейчас течёт?','Yes, it is flowing.','Да, она течёт.'), ('Do you heat water to make it hot?','Ты нагреваешь воду, чтобы сделать её горячей?','Yes, I heat water to make it hot.','Да, я нагреваю воду, чтобы сделать её горячей.'), ('What is the weather like in winter?','Какая погода зимой?','It is sometimes cool and rainy.','Иногда прохладно и дождливо.'), ('Do children go to school when there is a lot of snow?','Дети ходят в школу, когда много снега?','No, they do not. They stay home.','Нет. Они остаются дома.'), ]''' s=s[:start]+new+s[end:] s=s.replace("OUT='/tmp/workspace/esl-handbook/output/Present_Tenses_Conversation_Handbook.pdf'","OUT='/tmp/workspace/esl-handbook/output/Logical_Present_Tenses_Conversation_Handbook.pdf'") s=s.replace("'15 conversation questions and answers about water, weather, and changes.'","'15 connected conversation questions and answers about water, weather, and changes.'") open(p,'w',encoding='utf-8').write(s) PY python "/tmp/workspace/esl-handbook/src/create_handbook.py" && python - <<'PY' from pypdf import PdfReader p='/tmp/workspace/esl-handbook/output/Logical_Present_Tenses_Conversation_Handbook.pdf' r=PdfReader(p) print('pages:',len(r.pages),'size:',__import__('os').path.getsize(p)) t='\n'.join(x.extract_text() or '' for x in r.pages) for phrase in ['What happens to water on a very cold day?','Steam is a gas.','Do children go to school']: print(phrase, phrase in t) PY