violetcereza
commented
1 month ago Thanks again for building this! I have been using this for my project, but I wanted to intersperse different levels of headings with my TODO items. I had a lot of trouble figuring out context for my to-do items in the middle of a large notebook.
To accomplish this, I wrote a printing function that takes proj_outline() output and translates the tibble into something we can display with cli::tree(). Here's an example tree with to-do items interspersed:
test.qmd
└─Test
├─Quarto
├─Running Code
│ ├─TODO: fix this in the code
│ └─A sub header
│ └─TODO: here's a todo in the text
└─Back to header 1
├─Dont skip me
│ └─header 5
│ └─TODO: testing section
└─Another sub header
└─TODO: section testI am still playing with the formatting. I started using cli::format_inline() to incorporate the contextual links you designed, but it doesn't show up here.
Below is the code I'm using. First, I simplify the outline tibble to something that makes sense to me. Then, I make the indent variable which is based off of n_leading_hash. Then, I transform a tibble with "indents" into a tree structure for cli::tree(). I can rewrite this as a pull request eventually, but I want to take a shot at using lightparser and spend more time customizing the appearance with colors and stuff before I integrate into your codebase.
outline_data <- reuseme::file_outline(path = "test.qmd") %>%
# Convert the many is_ columns into mutually exclusive "outline row types"
pivot_longer(
names_to = "type", names_prefix = "is_", c(
starts_with("is_"), -is_md, -is_second_level_heading_or_more, -is_roxygen_comment, -is_notebook
)
) %>%
# # Double check that types are mututally exclusive
# filter(all(value == F), .by = c(file_short, title, line_id))
# filter(sum(value) > 1, .by = c(file_short, title, line_id))
filter(value == T) %>%
# We drop these because they don't serve to add much context to TODOs (they don't affect heirarchy)
filter(type != "tab_or_plot_title") %>%
# Some useful definitions!
mutate(
title = coalesce(outline_el, title_el),
file_short = fs::path_file(file),
n_leading_hash = type %>% case_match(
c("todo_fixme", "tab_or_plot_title") ~ NA,
.default = n_leading_hash
)
) %>%
# For each file, stick a item at the top of the outline
group_by(file, file_short) %>%
group_modify(\(data, group) data %>% add_row(
.before = 0,
n_leading_hash = -1,
title = group$file_short,
type = "file"
)) %>%
mutate(
# Processing how title displays based on type
print_title = type %>% case_match(
"todo_fixme" ~ link,
.default = link_rs_api
) %>% coalesce(title) %>% map_chr(cli::format_inline),
# Assign TODO items (and other items missing n_leading_hash)
# to be indented under the last seen header level
indent = coalesce(n_leading_hash, zoo::na.locf0(n_leading_hash+1)),
# If there are any headers that skip an intermediate level, pick them up
skip_level = indent > lag(indent)+1,
skip_level_should_be = ifelse(skip_level, lag(indent)+1, NA),
skip_level_adjust = case_when(
# All the items below on the outline should be adjusted backwards
skip_level ~ skip_level_should_be-indent,
# Unless we reach a point on the outline where we're back up in
# the hierachy, so stop adjusting.
indent <= zoo::na.locf0(skip_level_should_be) ~ 0
) %>%
# Carry the adjustments to later rows
zoo::na.locf0() %>% coalesce(0),
indent = indent + skip_level_adjust
) %>%
ungroup() %>%
select(title, type, n_leading_hash, indent, print_title)
outline_data %>%
mutate(
# Give items IDs so titles do not have to be unique
item_id = as.character(row_number()),
indent_wider = indent,
x = TRUE
) %>%
# We need these wide cumsum `header1` type fields to determine which items belong to which parents
pivot_wider(names_from = indent_wider, values_from = x, values_fill = F, names_prefix = "header") %>%
mutate(across(starts_with("header"), cumsum)) %>%
# For each row, pick the IDs of all direct children from the outline
pmap(function(...) with(list(..., childdata = .), tibble(
title,
print_title,
indent,
item_id,
type,
parent_level_id = get(stringr::str_c("header", indent)),
children_ids = childdata %>%
dplyr::rename(childindent = indent) %>%
dplyr::filter(
childindent == indent+1,
cumsum(childindent == indent) == parent_level_id
) %>%
dplyr::pull(item_id) %>% list()
))) %>%
list_rbind() %>%
# View()
select(item_id, children_ids, print_title) %>%
cli::tree()
And here's the qmd text I used:
---
title: "Test"
format: html
editor: visual
---
## Quarto
Quarto enables you to weave together content and executable code into a finished document. To learn more about Quarto see <https://quarto.org>.
## Running Code
When you click the **Render** button a document will be generated that includes both content and the output of embedded code. You can embed code like this:
```{r}
1 + 1
# TODO: fix this in the codeYou can add options to executable code like this
A sub header
#| echo: false
2 * 2The echo: false option disables the printing of code (only output is displayed).
TODO: here's a todo in the text
Back to header 1
Dont skip me
header 5
TODO: testing section
Another sub header
TODO: section test
olivroy
Please comment on this issue if you tried the
*_outline()functions and would like some functionality or encountered something unexpectedcf https://github.com/rstudio/rstudioapi/issues/153#issuecomment-2116139239
Current plans
roxygen2::parse_package()+ md (#29)Known issues