Files
social-media/scripts/generate-db-diagram.sh
Jonathan Bourdon 98c76a7d88
Some checks failed
deploy-socialize / image (push) Failing after 24s
deploy-socialize / deploy (push) Has been skipped
chore: group database diagram tables by module
2026-05-07 15:51:47 -04:00

483 lines
15 KiB
Bash
Executable File

#!/usr/bin/env bash
set -euo pipefail
SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
REPO_ROOT="$(cd "${SCRIPT_DIR}/.." && pwd)"
OUT_DIR="${DB_DIAGRAM_OUT_DIR:-${REPO_ROOT}/.artifacts/db-diagrams}"
mkdir -p "$OUT_DIR"
TABLES_TSV="${OUT_DIR}/tables.tsv"
RELATIONS_TSV="${OUT_DIR}/relations.tsv"
DOT_FILE="${OUT_DIR}/database-diagram.dot"
SVG_FILE="${OUT_DIR}/database-diagram.svg"
PNG_FILE="${OUT_DIR}/database-diagram.png"
HTML_FILE="${OUT_DIR}/database-diagram.html"
export PGCONNECT_TIMEOUT="${PGCONNECT_TIMEOUT:-5}"
export PGHOST="${PGHOST:-localhost}"
export PGPORT="${PGPORT:-5433}"
export PGDATABASE="${PGDATABASE:-socialize}"
export PGUSER="${PGUSER:-sa}"
PSQL_ARGS=(-v ON_ERROR_STOP=1 -X -A -t -F $'\t')
CONNECTION_LABEL="${PGHOST}:${PGPORT}/${PGDATABASE} as ${PGUSER}"
if [[ -n "${DATABASE_URL:-}" ]]; then
PSQL=(psql "${DATABASE_URL}" "${PSQL_ARGS[@]}")
CONNECTION_LABEL="DATABASE_URL"
elif docker ps --format '{{.Names}}' 2>/dev/null | grep -qx "socialize-postgres"; then
PSQL=(docker exec -i socialize-postgres psql -U "$PGUSER" -d "$PGDATABASE" "${PSQL_ARGS[@]}")
CONNECTION_LABEL="Docker container socialize-postgres/${PGDATABASE} as ${PGUSER}"
else
PSQL=(psql "${PSQL_ARGS[@]}")
fi
if [[ "${PSQL[0]}" == "psql" ]] && ! command -v psql >/dev/null 2>&1; then
echo "psql is required to generate the database diagram when not using the local Docker container." >&2
exit 1
fi
echo "Reading schema from ${CONNECTION_LABEL}..."
if ! "${PSQL[@]}" -c "select 1" >/dev/null 2>&1; then
echo "Could not connect to PostgreSQL." >&2
echo "Start infrastructure first, then provide credentials with DATABASE_URL, PGPASSWORD, or ~/.pgpass." >&2
exit 1
fi
"${PSQL[@]}" > "$TABLES_TSV" <<'SQL'
SELECT
c.table_schema || '.' || c.table_name AS table_id,
c.column_name,
CASE
WHEN c.data_type = 'character varying'
THEN 'varchar(' || c.character_maximum_length || ')'
WHEN c.data_type = 'character'
THEN 'char(' || c.character_maximum_length || ')'
WHEN c.data_type = 'text'
THEN 'text'
WHEN c.data_type = 'numeric' AND c.numeric_precision IS NOT NULL AND c.numeric_scale IS NOT NULL
THEN 'numeric(' || c.numeric_precision || ',' || c.numeric_scale || ')'
WHEN c.data_type = 'numeric' AND c.numeric_precision IS NOT NULL
THEN 'numeric(' || c.numeric_precision || ')'
WHEN c.data_type = 'timestamp with time zone'
THEN 'timestamptz'
WHEN c.data_type = 'timestamp without time zone'
THEN 'timestamp'
WHEN c.data_type = 'USER-DEFINED'
THEN c.udt_name
ELSE c.data_type
END AS formatted_type,
c.is_nullable,
concat_ws(' ',
CASE WHEN pk.column_name IS NULL THEN NULL ELSE 'PK' END,
CASE WHEN fk.column_name IS NULL THEN NULL ELSE 'FK' END,
CASE WHEN ix.column_name IS NULL THEN NULL ELSE 'IX' END
) AS column_markers
FROM information_schema.columns c
LEFT JOIN (
SELECT
kcu.table_schema,
kcu.table_name,
kcu.column_name
FROM information_schema.table_constraints tc
JOIN information_schema.key_column_usage kcu
ON kcu.constraint_catalog = tc.constraint_catalog
AND kcu.constraint_schema = tc.constraint_schema
AND kcu.constraint_name = tc.constraint_name
WHERE tc.constraint_type = 'PRIMARY KEY'
) pk
ON pk.table_schema = c.table_schema
AND pk.table_name = c.table_name
AND pk.column_name = c.column_name
LEFT JOIN (
SELECT DISTINCT
kcu.table_schema,
kcu.table_name,
kcu.column_name
FROM information_schema.table_constraints tc
JOIN information_schema.key_column_usage kcu
ON kcu.constraint_catalog = tc.constraint_catalog
AND kcu.constraint_schema = tc.constraint_schema
AND kcu.constraint_name = tc.constraint_name
WHERE tc.constraint_type = 'FOREIGN KEY'
) fk
ON fk.table_schema = c.table_schema
AND fk.table_name = c.table_name
AND fk.column_name = c.column_name
LEFT JOIN (
SELECT DISTINCT
indexed_schema.nspname AS table_schema,
indexed_table.relname AS table_name,
indexed_attribute.attname AS column_name
FROM pg_catalog.pg_index index_definition
JOIN pg_catalog.pg_class indexed_table
ON indexed_table.oid = index_definition.indrelid
JOIN pg_catalog.pg_namespace indexed_schema
ON indexed_schema.oid = indexed_table.relnamespace
JOIN pg_catalog.pg_attribute indexed_attribute
ON indexed_attribute.attrelid = indexed_table.oid
AND indexed_attribute.attnum = ANY(index_definition.indkey)
WHERE indexed_schema.nspname = 'public'
) ix
ON ix.table_schema = c.table_schema
AND ix.table_name = c.table_name
AND ix.column_name = c.column_name
WHERE c.table_schema = 'public'
AND c.table_name <> '__EFMigrationsHistory'
ORDER BY c.table_name, c.ordinal_position;
SQL
"${PSQL[@]}" > "$RELATIONS_TSV" <<'SQL'
SELECT
tc.constraint_name,
kcu.table_schema || '.' || kcu.table_name AS child_table,
kcu.column_name AS child_column,
ccu.table_schema || '.' || ccu.table_name AS parent_table,
ccu.column_name AS parent_column,
col.is_nullable,
CASE
WHEN EXISTS (
SELECT 1
FROM pg_catalog.pg_index unique_index
JOIN pg_catalog.pg_class unique_table
ON unique_table.oid = unique_index.indrelid
JOIN pg_catalog.pg_namespace unique_schema
ON unique_schema.oid = unique_table.relnamespace
JOIN pg_catalog.pg_attribute unique_attribute
ON unique_attribute.attrelid = unique_table.oid
AND unique_attribute.attnum = ANY(unique_index.indkey)
WHERE unique_schema.nspname = kcu.table_schema
AND unique_table.relname = kcu.table_name
AND unique_attribute.attname = kcu.column_name
AND unique_index.indisunique
AND unique_index.indnatts = 1
)
THEN 'YES'
ELSE 'NO'
END AS child_column_is_unique
FROM information_schema.table_constraints tc
JOIN information_schema.key_column_usage kcu
ON kcu.constraint_catalog = tc.constraint_catalog
AND kcu.constraint_schema = tc.constraint_schema
AND kcu.constraint_name = tc.constraint_name
JOIN information_schema.constraint_column_usage ccu
ON ccu.constraint_catalog = tc.constraint_catalog
AND ccu.constraint_schema = tc.constraint_schema
AND ccu.constraint_name = tc.constraint_name
JOIN information_schema.columns col
ON col.table_schema = kcu.table_schema
AND col.table_name = kcu.table_name
AND col.column_name = kcu.column_name
WHERE tc.constraint_type = 'FOREIGN KEY'
AND kcu.table_schema = 'public'
AND kcu.table_name <> '__EFMigrationsHistory'
AND ccu.table_schema = 'public'
AND ccu.table_name <> '__EFMigrationsHistory'
ORDER BY child_table, child_column, parent_table, parent_column;
SQL
html_escape() {
local value="$1"
value="${value//&/&amp;}"
value="${value//</&lt;}"
value="${value//>/&gt;}"
value="${value//\"/&quot;}"
printf '%s' "$value"
}
node_id() {
local value="$1"
value="${value//./__}"
value="${value//-/__}"
printf '%s' "$value"
}
dot_label_escape() {
local value="$1"
value="${value//\\/\\\\}"
value="${value//\"/\\\"}"
printf '%s' "$value"
}
table_module() {
local table_name="${1#public.}"
case "$table_name" in
AspNet*) printf 'Identity' ;;
OrganizationMemberships | Organizations) printf 'Organizations' ;;
Workspaces | WorkspaceInvites) printf 'Workspaces' ;;
Clients) printf 'Clients' ;;
Campaigns) printf 'Campaigns' ;;
Channels) printf 'Channels' ;;
ContentItems | ContentItemRevisions | ContentItemActivityEntries) printf 'ContentItems' ;;
Assets | AssetRevisions) printf 'Assets' ;;
Comments) printf 'Comments' ;;
ApprovalDecisions | ApprovalRequests | ApprovalWorkflowInstances | WorkspaceApprovalStepConfigurations) printf 'Approvals' ;;
NotificationEvents) printf 'Notifications' ;;
Feedback*) printf 'Feedback' ;;
CalendarCatalogEntries | CalendarEvents | CalendarSources | UserCalendarExportFeeds) printf 'CalendarIntegrations' ;;
*) printf 'Other' ;;
esac
}
cluster_id() {
local value="$1"
value="${value//[^[:alnum:]]/_}"
printf 'cluster_%s' "$value"
}
write_table_node() {
local table_name="$1"
local id title
id="$(node_id "$table_name")"
title="$(html_escape "${table_name#public.}")"
printf ' %s [label=<\n' "$id"
printf ' <TABLE BORDER="1" CELLBORDER="1" CELLSPACING="0" CELLPADDING="6" COLOR="#94a3b8">\n'
printf ' <TR><TD BGCOLOR="#0f766e" COLSPAN="3"><FONT COLOR="white"><B>%s</B></FONT></TD></TR>\n' "$title"
printf '%s' "${TABLE_COLUMNS[$table_name]}"
printf ' </TABLE>\n'
printf ' >];\n\n'
}
declare -A TABLE_COLUMNS
declare -A TABLE_SEEN
declare -A MODULE_SEEN
declare -A MODULE_TABLES
declare -a TABLE_ORDER
declare -a MODULE_ORDER
while IFS=$'\t' read -r table_name column_name column_type nullable column_markers; do
[[ -n "$table_name" ]] || continue
if [[ -z "${TABLE_SEEN[$table_name]:-}" ]]; then
TABLE_SEEN["$table_name"]=1
TABLE_ORDER+=("$table_name")
module_name="$(table_module "$table_name")"
if [[ -z "${MODULE_SEEN[$module_name]:-}" ]]; then
MODULE_SEEN["$module_name"]=1
MODULE_ORDER+=("$module_name")
fi
MODULE_TABLES["$module_name"]+="${table_name}"$'\n'
fi
label="$(html_escape "$column_name")"
type_label="$(html_escape "$column_type")"
marker_label="$(html_escape "$column_markers")"
[[ -n "$marker_label" ]] || marker_label="&#160;"
nullable_label=""
[[ "$nullable" == "YES" ]] && nullable_label=' <FONT POINT-SIZE="10">nullable</FONT>'
TABLE_COLUMNS["$table_name"]+=$' <TR><TD ALIGN="LEFT"><FONT POINT-SIZE="10">'"${marker_label}"$'</FONT></TD><TD ALIGN="LEFT" PORT="'$(html_escape "$column_name")$'"><FONT FACE="monospace">'"${label}"$'</FONT></TD><TD ALIGN="LEFT"><FONT POINT-SIZE="10">'"${type_label}${nullable_label}"$'</FONT></TD></TR>\n'
done < "$TABLES_TSV"
{
cat <<'DOT'
digraph database {
graph [
rankdir=LR,
bgcolor="white",
pad=0.4,
nodesep=0.6,
ranksep=1.0,
splines=spline,
overlap=false
];
node [
shape=plain,
fontname="Arial"
];
edge [
color="#64748b",
arrowsize=0.7,
fontname="Arial",
fontsize=10,
labeldistance=1.8,
labelangle=22
];
legend [label=<
<TABLE BORDER="1" CELLBORDER="0" CELLSPACING="0" CELLPADDING="6" COLOR="#94a3b8">
<TR><TD BGCOLOR="#334155" COLSPAN="2"><FONT COLOR="white"><B>Legend</B></FONT></TD></TR>
<TR><TD ALIGN="LEFT">PK</TD><TD ALIGN="LEFT">primary key column</TD></TR>
<TR><TD ALIGN="LEFT">FK</TD><TD ALIGN="LEFT">foreign key column</TD></TR>
<TR><TD ALIGN="LEFT">IX</TD><TD ALIGN="LEFT">indexed column</TD></TR>
<TR><TD ALIGN="LEFT">solid line</TD><TD ALIGN="LEFT">required FK</TD></TR>
<TR><TD ALIGN="LEFT">dashed line</TD><TD ALIGN="LEFT">nullable FK</TD></TR>
<TR><TD ALIGN="LEFT">1</TD><TD ALIGN="LEFT">exactly one referenced row</TD></TR>
<TR><TD ALIGN="LEFT">0..1</TD><TD ALIGN="LEFT">zero or one referencing row</TD></TR>
<TR><TD ALIGN="LEFT">0..*</TD><TD ALIGN="LEFT">zero or many referencing rows</TD></TR>
<TR><TD ALIGN="LEFT" COLSPAN="2">Edges point from referenced table to referencing table.</TD></TR>
</TABLE>
>];
DOT
for module_name in "${MODULE_ORDER[@]}"; do
id="$(cluster_id "$module_name")"
label="$(html_escape "$module_name")"
printf ' subgraph %s {\n' "$id"
printf ' label=< <B>%s</B> >;\n' "$label"
printf ' color="#cbd5e1";\n'
printf ' penwidth=1.5;\n'
printf ' style="rounded";\n'
printf ' margin=18;\n'
printf ' fontname="Arial";\n'
printf ' fontsize=16;\n'
printf ' fontcolor="#334155";\n\n'
while IFS= read -r table_name; do
[[ -n "$table_name" ]] || continue
write_table_node "$table_name"
done <<< "${MODULE_TABLES[$module_name]}"
printf ' }\n\n'
done
while IFS=$'\t' read -r constraint_name child_table child_column parent_table parent_column nullable child_column_is_unique; do
[[ -n "$constraint_name" ]] || continue
child_id="$(node_id "$child_table")"
parent_id="$(node_id "$parent_table")"
child_cardinality="0..*"
[[ "$child_column_is_unique" == "YES" ]] && child_cardinality="0..1"
label="$(dot_label_escape "${parent_column} -> ${child_column}")"
style="solid"
[[ "$nullable" == "YES" ]] && style="dashed"
printf ' %s -> %s [label="%s", taillabel="1", headlabel="%s", style="%s"];\n' "$parent_id" "$child_id" "$label" "$child_cardinality" "$style"
done < "$RELATIONS_TSV"
printf '}\n'
} > "$DOT_FILE"
if command -v dot >/dev/null 2>&1; then
dot -Tsvg "$DOT_FILE" -o "$SVG_FILE"
dot -Tpng "$DOT_FILE" -o "$PNG_FILE"
else
echo "Graphviz dot was not found; wrote only ${DOT_FILE}." >&2
fi
cat > "$HTML_FILE" <<HTML
<!doctype html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>Socialize Database Diagram</title>
<style>
html,
body {
margin: 0;
min-height: 100%;
font-family: Arial, sans-serif;
background: #f8fafc;
color: #0f172a;
}
header {
position: sticky;
top: 0;
z-index: 1;
padding: 12px 16px;
border-bottom: 1px solid #cbd5e1;
background: #ffffff;
}
.bar {
display: flex;
gap: 16px;
align-items: center;
margin-bottom: 8px;
}
header h1 {
margin: 0;
font-size: 18px;
line-height: 1.3;
}
header a {
color: #0f766e;
font-size: 14px;
text-decoration: none;
}
.legend {
display: flex;
flex-wrap: wrap;
gap: 8px 16px;
margin: 0;
padding: 0;
font-size: 13px;
color: #334155;
list-style: none;
}
.line {
display: inline-block;
width: 28px;
margin-right: 6px;
border-top: 2px solid #64748b;
vertical-align: middle;
}
.line.dashed {
border-top-style: dashed;
}
main {
padding: 16px;
}
.canvas {
width: max-content;
min-width: calc(100vw - 32px);
padding: 16px;
border: 1px solid #cbd5e1;
background: #ffffff;
}
object {
display: block;
min-width: 100%;
min-height: calc(100vh - 104px);
}
</style>
</head>
<body>
<header>
<div class="bar">
<h1>Socialize Database Diagram</h1>
<a href="./database-diagram.svg">SVG</a>
<a href="./database-diagram.png">PNG</a>
<a href="./database-diagram.dot">DOT</a>
</div>
<ul class="legend">
<li><span class="line"></span>required FK</li>
<li><span class="line dashed"></span>nullable FK</li>
<li><strong>1</strong>: exactly one referenced row</li>
<li><strong>0..1</strong>: zero or one referencing row</li>
<li><strong>0..*</strong>: zero or many referencing rows</li>
<li>Edges point from referenced table to referencing table.</li>
</ul>
</header>
<main>
<div class="canvas">
<object data="./database-diagram.svg" type="image/svg+xml">
Open database-diagram.svg or database-diagram.png from this folder.
</object>
</div>
</main>
</body>
</html>
HTML
echo "Wrote:"
echo " ${HTML_FILE}"
[[ -f "$SVG_FILE" ]] && echo " ${SVG_FILE}"
[[ -f "$PNG_FILE" ]] && echo " ${PNG_FILE}"
echo " ${DOT_FILE}"