wherewasi ========= [13.6.2018] units and storage * improved text preprocessing ** introducing units ** remove text in brackets, as it seems to be added information but mostly not needed * adding mongo db storage to persist recipes * TODO: extract ingredients * TODO: extract graph with chained prod = function(prod, ...) objects * TODO: start to parse multiple sample pages [23.5.2018] language detection and ingredients * preprocessing input, detecting language, english support only for now * creating ingredient list [17.5.2018] removing appengine * simplifying setup (for now), without google appengine * splitting segments * adding groups (rendering) [11.5.2018] process fetched snipped and display * apply simple cleanup * render using tag cloud like design [6.4.2018] Fetch and send... * created a bookmarklet to fetch selected html * send html code to backend service [27.2.2018] The vision for first steps... * use gcloud to host a python webservice * input: ** structured json recipe ** url to webpage with content to be extracted * output: render recipe to coocook graph * flask / html rendering install ======= * run python service on localhost (see testing in section below) * install bookmarklet * select recipe from a website and send content to coocook renderer build/test/run ============== ##virtualenv virtualenv env source env/bin/activate ## eventually needed for spacy language support export LC_ALL=en_US.UTF-8 export LANG=en_US.UTF-8 ##requirements pip install -t lib -r requirements.txt ##testing ./run.sh -> http://localhost:5000/info to see the results. Requirements ============ - Python 3.6 - Multiple python libraries described in requirements.txt - Spacy Models pip install -U spacy python -m spacy download en python -m spacy download de documentation/reads =================== * spaCy: https://nlpforhackers.io/complete-guide-to-spacy/