Details of the implementation of the gEDA/gschem plugin

The plugin for parsing and printing schematic files in the gEDA/gschem format is at ^[1] Github branch.
Get it by cloning the git repo : git clone git@github.sksavant/gnucap.git –branch lang_gschem or get this zip file : zipfile

If you already have latest snapshot of gnucap installed you just need to get the following files: 'src/lang_gschem.cc', 'src/d_net.cc', 'src/d_place.cc'. Also some change is need in lang_verilog.cc to print verilog code correctly. Some examples are present in 'geda/examples/ '

After getting the files compile the plugin with the following command:

  g++ -DHAVE_GEDA -shared -fPIC -o lang_geda.so lang_geda.cc d_net.cc d_place.cc  $(pkg-config --cflags gtk+-2.0 --libs libgeda)

And load it after running gnucap as follows:

  gnucap>load ./lang_geda.so 

This line can be added to '~/.gnucaprc' to load the plugin every time gnucap starts.
Switch the language to gEDA/gschem with the following command:

  gnucap>option lang=gschem OR
  gnucap>gschem

When the command 'gschem' is given, CMD_GSCHEM::do_it is called. After setting the language to gschem, parse_top_item is called, which will call new__instance after getting a line from the command stream.
Now depending on the type of the line parsed, which is found from find_type_in_string, corresponding parse_.. function is called. If a command or dev_comment is got, parse_command or parse_comment is called respectively.

If component is found, then CMD_C::do_it is called, which

• creates a new empty placeholder model (MODEL_SUBCKT)
• calls parse_componmod which will
  • get the symbol file name
  • parse the symbol file and set the ports by name
  • resets the cmd
  • returns NULL if the model is not needed i.e, it is graphical
• If the model is not graphical ,push back into the CARD_LIST
• set componentname as the label of the model and componentmod as false and call new__instance.

If net or place or any other is got, parse_instance is called which will call respective parse_ functions.
When parse_net is called it will :

• Get the x y coordinates of the two ports
• parse the symbol file
• Call create_place which will create a CS with the place parameters and parse it through new__instance.
• Check if there are any attributes after this line, parse them into set_param_by_name '{..}

If a component is got, call parse_component which will :

• Get the position, angle, mirror flag and the symbol file name of the component
• Parse the symbol file associated with the component to get the ports and get the ports' coordinates relative to the component.
• Depending on the angle and mirror flag, get the absolute position of the ports of the component, create a new place with the position of nodes and call set_port_by_index to set the ports
• Check if there are any attributes after this line, parse them into set_param_by_name '{..}