README update

Author: diije

README.md

@@ -4,29 +4,33 @@ A simple Python client example to send requests and get results from [DataForSEO
 
 ## Setup
 
-1. Clone this repo:  
+__1. Clone this repo:__  
 
-    $ git clone git@gitlab.com:databulle/dataforseo-serp-api-python-client.git  
+    $ git clone git@gitlab.com:databulle/dataforseo-serp-api-python-client.git
     $ cd dataforseo-serp-api-python-client  
 
-2. (Optional) Setup a new [pyenv](https://github.com/pyenv/pyenv) virtual environment:  
 
-    $ pyenv virtualenv 3.7.0 dataforseo  
+__2. (Optional) Setup a new [pyenv](https://github.com/pyenv/pyenv) virtual environment:__  
+
+    $ pyenv virtualenv 3.7.0 dataforseo
     $ pyenv local dataforseo  
 
-3. Get your credentials from DataForSEO:  
+
+__3. Get your credentials from DataForSEO:__  
 
 Go to [the API dashboard](https://app.dataforseo.com/api-dashboard) and get your `API Login` and `API Password`.  
 
-4. Create you `config.ini` file:  
+
+__4. Create you `config.ini` file:__  
 
 Copy the sample config file:  
 
     $ cp config.sample.ini config.ini  
 
 Edit the new file and fill in your login and password  
 
-5. That's all!  
+
+__5. That's all!__  
 
 ### Sandbox  
 
@@ -46,6 +50,26 @@ Remember to switch it back when things get real ;)
 
 ### 1. Post tasks
 
+Put the list of keywords in a text file (one request per line), then run:  
+
+    $ python tasks_post.py --input yourfile.txt  
+
+The script will send the keywords to the API by batches of 100 requests, and write a report containing the `task_id` for each keyword.  
+
+The `input` argument is required. Other options are available:  
+- `config`: configuration file to use (default: `config.ini`)  
+- `input`: input file with the list of requests (required)  
+- `output`: output basename for the report (default: `keyword-requests`)  
+- `language_code`: language code for the requests (default: `fr`)  
+- `location_code`: location code for requests (default: "2250" for France, get other codes on <https://api.dataforseo.com/v3/serp/google/locations>)  
+- `nb_results`: number of results to ask for (default: `10`)  
+- `device`: choice between `desktop` (default) and `mobile`  
+- `priority`: choose the priority queue between `low` (default) and `high` (note: `high` priority requests will be charged more by DataForSEO)  
+- `batch`: number of requests to send for each batch, between 1 and 100 (default: `100`)  
+- `delay`: delay in seconds between batches of requests (default: `10`)  
+- `sep`: CSV separator for `output`(default: `;`)  
+
+
 ### 2. See if tasks are ready
 
 This step is not required, but will show you how many of your requests have been processed and are available for download.